Bug 1468436
| Summary: | Libreoffice Writer crashing with segmentation fault in libjvm.so _expand_stack_to when wiki plugin installed | ||||||
|---|---|---|---|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Yura <yuriykuzin> | ||||
| Component: | kernel | Assignee: | Kernel Maintainer List <kernel-maint> | ||||
| Status: | CLOSED INSUFFICIENT_DATA | QA Contact: | Fedora Extras Quality Assurance <extras-qa> | ||||
| Severity: | medium | Docs Contact: | |||||
| Priority: | unspecified | ||||||
| Version: | 30 | CC: | ableisch, adam, ahughes, airlied, antonio.montagnani, biswas.pinaki1708, bskeggs, caolanm, dbhole, dhgutteridge, dtardon, dwrobel, eduard.vopicka, emelenas, erack, ewk, extras-qa, hdegoede, ichavero, itamar, jarodwilson, jerboaa, jglisse, john.j5live, jonathan, josef, jvanek, kernel-maint, labbott, linville, lukebenes, mchehab, mjg59, moosedrool42, msrb, mvala, omajid, romieu, samuel-rhbugs, sbergman, steved, steven.scheider, tsiraut, yuriykuzin | ||||
| Target Milestone: | --- | Keywords: | Reopened | ||||
| Target Release: | --- | ||||||
| Hardware: | i686 | ||||||
| OS: | Linux | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | Doc Type: | If docs needed, set a value | |||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2019-09-17 20:07:41 UTC | Type: | Bug | ||||
| Regression: | --- | Mount Type: | --- | ||||
| Documentation: | --- | CRM: | |||||
| Verified Versions: | Category: | --- | |||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||
| Embargoed: | |||||||
| Bug Depends On: | |||||||
| Bug Blocks: | 1489998 | ||||||
| Attachments: |
|
||||||
I don't get a crash under f25 on installing the publisher and checking tools->options->internet->mediawiki so I can't reproduce this --backtrace unfortunately will stop at the first SEGV and java generates loads of them so its tricky to get the right backtrace automatically caolanm->sberg: any ideas ? it's possible that this _expand_stack_to isn't the usual harmless JVM SIGSEGV but something different. _rene_ reported seeing this on Debian: https://buildd.debian.org/status/fetch.php?pkg=libreoffice&arch=i386&ver=1%3A5.3.4-1&stamp=1498442560&raw=0 hum, maybe we have hit a kernel bug then ? https://lkml.org/lkml/2017/7/4/610 https://lwn.net/Articles/727206/ anyhow, fix is presumably required either in openjdk or the kernel and not in libreoffice so I'll push it down a level Hi sorry for an intrusion just update I've tried with Fedora 25 installed on virtualbox and it's reproduced some info ---------- Linux localhost.localdomain 4.11.8-200.fc25.i686+PAE #1 SMP Thu Jun 29 16:36:01 UTC 2017 i686 i686 i386 GNU/Linux ---------- libreoffice.i686 1:5.2.7.2-4.fc25 ---------- libreoffice-wiki-publisher.i686 1:5.2.7.2-4.fc25 ---------- java-1.8.0-openjdk.i686 1:1.8.0.131-1.b12.fc25 tell me if I can provide more info Are you able to verify that using an older kernel does _not_ cause this crash? (In reply to Deepak Bhole from comment #6) > Are you able to verify that using an older kernel does _not_ cause this > crash? yes lucky me the last saved kernel work, nice idea and looks like problem localized simply I've boot into old kernel and no crash. so results: ---- NO CRASH ---- Linux happyfedora 4.11.5-100.fc24.i686+PAE #1 SMP Wed Jun 14 17:44:34 UTC 2017 i686 i686 i386 GNU/Linux ------------------ ---- CRASH ---- Linux happyfedora 4.11.6-101.fc24.i686+PAE #1 SMP Tue Jun 20 16:50:52 UTC 2017 i686 i686 i386 GNU/Linux --------------- *** Bug 1473485 has been marked as a duplicate of this bug. *** *** Bug 1473920 has been marked as a duplicate of this bug. *** I'm hitting the same bug; in my case the issue is caused by the LanguageTool plugin for Writer, which requires Java. If I disable LanguageTool, Writer runs fine. This is again on i686. I cannot reproduce the problem on x86_64. I can also reproduce this with 4.12 kernels, e.g. kernel-4.12.0-1.fc27. I updated to the latest OpenJDK in updates-testing (java-1.8.0-openjdk-1.8.0.141-1.b16), and that makes no difference for me. This message is a reminder that Fedora 24 is nearing its end of life. Approximately 2 (two) weeks from now Fedora will stop maintaining and issuing updates for Fedora 24. It is Fedora's policy to close all bug reports from releases that are no longer maintained. At that time this bug will be closed as EOL if it remains open with a Fedora 'version' of '24'. Package Maintainer: If you wish for this bug to remain open because you plan to fix it in a currently maintained version, simply change the 'version' to a later Fedora version. Thank you for reporting this issue and we are sorry that we were not able to fix it before Fedora 24 is end of life. If you would still like to see this bug fixed and are able to reproduce it against a later version of Fedora, you are encouraged change the 'version' to a later Fedora version prior this bug is closed as described in the policy above. Although we aim to fix as many bugs as possible during every release's lifetime, sometimes those efforts are overtaken by events. Often a more recent Fedora release includes newer upstream software that fixes bugs or makes them obsolete. Same issue with Fedora 26 and *all* available i686 kernel for FC26.
java-1.8.0-openjdk-headless-1.8.0.131-7.b12.fc26.i686
libreoffice-core-5.3.4.2-4.fc26.i686
Workaround: install kernel-PAE-4.8.6-300.fc25.i686 from fedora-25
# dnf install kernel-PAE-4.8.6-300.fc25 --releasever=25
# journalctl -xn
Jul 26 19:56:39 example systemd-coredump[2359]: Process 2346 (soffice.bin) of user 1000 dumped core.
Stack trace of thread 2346:
#0 0x00000000a789e0a5 _ZL16_expand_stack_toPh (libjvm.so)
#1 0x00000000a78a0b6b _ZN2os5Linux21manually_expand_stackEP10JavaThreadPh (libjvm.so)
#2 0x00000000a78ad147 _ZN2os18create_main_threadEP10JavaThread (libjvm.so)
#3 0x00000000a7a11525 _ZN7Threads9create_vmEP14JavaVMInitArgsPb (libjvm.so)
#4 0x00000000a766b185 JNI_CreateJavaVM (libjvm.so)
#5 0x00000000b156266e _Z34jfw_plugin_startJavaVirtualMachinePK8JavaInfoPK12JavaVMOptionlPP7JavaVM_PP7JNIEnv_ (libjvmfwklo.so)
#6 0x00000000b15745fd _Z11jfw_startVMPK8JavaInfoP12JavaVMOptionlPP7JavaVM_PP7JNIEnv_ (libjvmfwklo.so)
#7 0x00000000a811d60a _ZN11stoc_javavm18JavaVirtualMachine9getJavaVMERKN3com3sun4star3uno8SequenceIaEE (libjavavmlo.so)
#8 0x00000000a8f047e8 _ZN15stoc_javaloader19JavaComponentLoader13getJavaLoaderEv (libjavaloaderlo.so)
#9 0x00000000a8f061b7 _ZN15stoc_javaloader19JavaComponentLoader17writeRegistryInfoERKN3com3sun4star3uno9ReferenceINS3_8registry12XRegistryKeyEEERKN3rtl8OUStringE
#10 0x00000000a814d72d _ZN12_GLOBAL__N_126ImplementationRegistration10doRegisterERKN3com3sun4star3uno9ReferenceINS3_4lang22XMultiComponentFactoryEEERKNS5_INS4_17X
#11 0x00000000a814fc1e _ZN12_GLOBAL__N_126ImplementationRegistration15prepareRegisterERKN3rtl8OUStringES4_S4_RKN3com3sun4star3uno9ReferenceINS7_8registry15XSimple
#12 0x00000000a9d7c313 _ZN11dp_registry7backend9component12_GLOBAL__N_111BackendImpl20ComponentPackageImpl15processPackage_ERN3osl15ResettableGuardINS5_5MutexEEEb
#13 0x00000000a9d9decf _ZN11dp_registry7backend7Package19processPackage_implEbbRKN3com3sun4star3uno9ReferenceINS4_4task13XAbortChannelEEERKNS6_INS4_3ucb19XCommand
#14 0x00000000a9d9e6ff _ZN11dp_registry7backend7Package15registerPackageEhRKN3com3sun4star3uno9ReferenceINS4_4task13XAbortChannelEEERKNS6_INS4_3ucb19XCommandEnvir
#15 0x00000000a9dc85cb _ZN11dp_registry7backend6bundle12_GLOBAL__N_111BackendImpl11PackageImpl15processPackage_ERN3osl15ResettableGuardINS5_5MutexEEEbbRKN3rtl9Ref
#16 0x00000000a9d9decf _ZN11dp_registry7backend7Package19processPackage_implEbbRKN3com3sun4star3uno9ReferenceINS4_4task13XAbortChannelEEERKNS6_INS4_3ucb19XCommand
#17 0x00000000a9d9e6ff _ZN11dp_registry7backend7Package15registerPackageEhRKN3com3sun4star3uno9ReferenceINS4_4task13XAbortChannelEEERKNS6_INS4_3ucb19XCommandEnvir
#18 0x00000000a9d4a846 _ZN10dp_manager16ExtensionManager17activateExtensionERKN3com3sun4star3uno8SequenceINS4_9ReferenceINS3_10deployment8XPackageEEEEEbbRKNS6_INS
#19 0x00000000a9d4cea2 _ZN10dp_manager16ExtensionManager11synchronizeERKN3com3sun4star3uno9ReferenceINS3_4task13XAbortChannelEEERKNS5_INS3_3ucb19XCommandEnvironme
#20 0x00000000b672127a _ZN7dp_misc16syncRepositoriesEbRKN3com3sun4star3uno9ReferenceINS2_3ucb19XCommandEnvironmentEEE (libdeploymentmisclo.so)
#21 0x00000000b75ee7d5 _ZN7desktop7Desktop32SynchronizeExtensionRepositoriesEv (libsofficeapp.so)
#22 0x00000000b75e0802 _ZN7desktop7Desktop4MainEv (libsofficeapp.so)
#23 0x00000000b4230222 _Z10ImplSVMainv (libvcllo.so)
#24 0x00000000b4230359 _Z6SVMainv (libvcllo.so)
#25 0x00000000b7616d35 soffice_main (libsofficeapp.so)
#26 0x000000008001f5cc main (soffice.bin)
#27 0x00000000b73d32f3 __libc_start_main (libc.so.6)
#28 0x000000008001f60f _start (soffice.bin)
I was getting a similar crash, but no wiki plugin installed. But, given the idea from this thread, I removed libreoffice-writer2latex-1.0.2-20.fc26.i686 and bingo - working again. Let me know if more diagnostic information is required/helpful - but I expect you can reproduce. Seems pretty decisive. Uninstall = working. Install = crashing. Hi, is this problem still happening? One of the more recent kernel updates should have addressed this. My post, yesterday, was on a freshly upgraded/patched Fedora 26, running 4.11.11-300.fc26.i686. (In reply to Deepak Bhole from comment #14) > Hi, is this problem still happening? One of the more recent kernel updates > should have addressed this. Hi, yes still happening after upgrade to F26 Updated components versions now: ----------------- # cat /etc/fedora-release Fedora release 26 (Twenty Six) ----------------- ----------------- # uname -a Linux happyfedora 4.11.8-300.fc26.i686+PAE #1 SMP Thu Jun 29 20:38:21 UTC 2017 i686 i686 i386 GNU/Linux ----------------- ----------------- # dnf list libreoffice libreoffice.i686 1:5.3.4.2-6.fc26 ----------------- ----------------- # dnf list libreoffice-wiki-publisher libreoffice-wiki-publisher.i686 1:5.3.4.2-6.fc26 ----------------- ----------------- # dnf list java java-1.8.0-openjdk.i686 1:1.8.0.141-1.b16.fc26 ----------------- (In reply to Yura from comment #16) > (In reply to Deepak Bhole from comment #14) > > Hi, is this problem still happening? One of the more recent kernel updates > > should have addressed this. > > Hi, yes still happening after upgrade to F26 > > Updated components versions now: > ----------------- > # cat /etc/fedora-release > Fedora release 26 (Twenty Six) > ----------------- > > ----------------- > # uname -a > Linux happyfedora 4.11.8-300.fc26.i686+PAE #1 SMP Thu Jun 29 20:38:21 UTC > 2017 i686 i686 i386 GNU/Linux > ----------------- > This is a much older kernel. Are you able to try with the latest? Latest stable in F26 is 4.11.12: https://bodhi.fedoraproject.org/updates/?packages=kernel (In reply to Deepak Bhole from comment #17) > This is a much older kernel. Are you able to try with the latest? Latest > stable in F26 is 4.11.12: > https://bodhi.fedoraproject.org/updates/?packages=kernel hm currently it's not available for me in repos for FC26 but before upgrade from FC24 I have it, so I'm able to boot into FC24 4.11.12 on upgraded FC26 kernel-PAE-4.11.12-100.fc24.i686 wiki still crashed my current FC26 kernel is kernel-PAE-4.11.11-300.fc26.i686 here it's too crashing This is still an issue with kernel 4.12.6-300.fc26.i686, built in Koji but not yet in updates-testing. *** Bug 1479762 has been marked as a duplicate of this bug. *** *** Bug 1483537 has been marked as a duplicate of this bug. *** *** Bug 1483537 has been marked as a duplicate of this bug. *** There is a workaround for this issue, for users who really need Java enabled with LibreOffice. Add the kernel parameter "stack_guard_gap=1". I tested this and found it works for me. (The source for this information is https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=865303#230.) Still otherwise an issue with kernel 4.12.9-300.fc26.i686. *** Bug 1497523 has been marked as a duplicate of this bug. *** Same thing that comment #23 with java-1.8.0-openjdk-1.8.0.151-1.b12.fc26.i386 and any of - kernel-4.12.14-300.fc26.i686 - kernel-4.13.11-200.fc26.i686 -- Ueimor (In reply to Francois Romieu from comment #25) > Same thing that comment #23 with > java-1.8.0-openjdk-1.8.0.151-1.b12.fc26.i386 > and any of > - kernel-4.12.14-300.fc26.i686 > - kernel-4.13.11-200.fc26.i686 > > -- > Ueimor Exactly, I confirm comment #23 and #25. Thanks David H. Gutteridge for the solution. *** Bug 1471293 has been marked as a duplicate of this bug. *** Is there a chance to get this fixed? This issue now started to hit me when doing LibreOffice Flatpak builds on Flathub (with a Flatpak-provided OpenJDK 9), which includes doing i386 builds, which includes running certain LibreOffice tests during the build that instantiate a JVM in the test process. These tests crash with SIGSEGV in _expand_stack_to, on my local machine with kernel 4.14.16-300.fc27.x86_64 (where I could debug the Flatpak build failure down to be this issue) as well as on various Flathub builders like flathub-builder-pdx1 with kernel 3.10.0-693.11.6.el7.x86_64 (where I very much assume the Flatpak build failures to be caused by this issue given the symptoms in the build logs; see also <https://cgit.freedesktop.org/libreoffice/core/commit/?id=9cf2616c5e709b595eeee6ab88dacdfad2003f98> "Work around i386 kernel vs. JVM bug for now by disabling all tests on i386"). (Debian seems to have addressed this with a kernel patch, see <https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=865303#492>.) This is a problem in the kernel and unfortunately there is nothing the JVM can do. RHEL kernel was fixed with this update: https://access.redhat.com/errata/RHSA-2017:2412 Unfortunately I don't know enough about the Fedora kernel to identify the exact fix. Here is the relevant portion that may help the Fedora kernel maintainers: "When upgrading to kernel with the fix for stack guard flaw, a crash could occur in Java Virtual Machine (JVM) environments, which attempted to implement their own stack guard page. With this update, the underlying source code has been fixed to consider the PROT_NONE mapping as a part of the stack, and the crash in JVM no longer occurs under the described circumstances. (BZ#1467938)" Re-assigning to kernel. This needs to be tracked as an i686 issue. There was never a consensus upstream about this. When it hanged state, reaction to <Ctrl>+<Alt>+<Del> is "reboot: Restarting system" followed by immediate reboot. Sorry for my above comment, my mistake, it is not related to this problem. This message is a reminder that Fedora 26 is nearing its end of life. Approximately 4 (four) weeks from now Fedora will stop maintaining and issuing updates for Fedora 26. It is Fedora's policy to close all bug reports from releases that are no longer maintained. At that time this bug will be closed as EOL if it remains open with a Fedora 'version' of '26'. Package Maintainer: If you wish for this bug to remain open because you plan to fix it in a currently maintained version, simply change the 'version' to a later Fedora version. Thank you for reporting this issue and we are sorry that we were not able to fix it before Fedora 26 is end of life. If you would still like to see this bug fixed and are able to reproduce it against a later version of Fedora, you are encouraged change the 'version' to a later Fedora version prior this bug is closed as described in the policy above. Although we aim to fix as many bugs as possible during every release's lifetime, sometimes those efforts are overtaken by events. Often a more recent Fedora release includes newer upstream software that fixes bugs or makes them obsolete. This is still relevant for Fedora 27 (and probably 28, though I'm not in a position to confirm that at the moment). Fedora 26 changed to end-of-life (EOL) status on 2018-05-29. Fedora 26 is no longer maintained, which means that it will not receive any further security or bug fix updates. As a result we are closing this bug. If you can reproduce this bug against a currently maintained version of Fedora please feel free to reopen this bug against that version. If you are unable to reopen this bug, please file a new report against the current release. If you experience problems, please add a comment to this bug. Thank you for reporting this bug and we are sorry it could not be fixed. I could reproduce this in Fedora 27 and still can reproduce it in Fedora 28. How do I request that this bug be reopened? You just did. Why are you still using a 32-bit kernel? If you aren't, then it's not this bug. [root@localhost ~]# grep sizes /proc/cpuinfo address sizes : 36 bits physical, 32 bits virtual [root@localhost ~]# How should I run 64bit kernel on this? :-) What kind of computer are you using that doesn't have 64-bit? I doubt there is much interest in trying to fix this since it's 32-bit only and that is being deprecated. There is a simple workaround provided previously. Just add that kernel option to the command line in grub. Hi Samuel.
Base Board Information
Manufacturer: Intel Corporation
Product Name: D845EPT2
processor : 0
vendor_id : GenuineIntel
cpu family : 15
model : 2
model name : Intel(R) Pentium(R) 4 CPU 2.40GHz
This machine in continuous service since approx 2002, runs w/o X for most of the time and perfectly fulfills its assigned role. With exception of fans and AGP graphic card, nothing broke during all these years.
Provided that there are two PCs with exactly the same HW configuration as spares and these are being powered-up each month for 8 hours to keep capacitors etc in good condition, I am too lazy to replace this.
Fortunately, I do not have to pay the electricity consumed by this machine. Should I do so, the machine will be immediately replaced by the Raspberry or something like that in appropriate configuration.
But once it happened that I had to edit some local file in a hurry, so I started X and libreoffice and it did not work. That is all.
Re "I doubt there is much interest in trying to fix this since it's 32-bit only and that is being deprecated." I can remember the Tesla 200 mainframe at the Czech Technical University. Only few days before it has been definitely turned off and moved to scrap overnight by its technical staff, full periodic profylactic work has been done by the same staff and everything repaired and tuned to the best operational parameters. At that moment, that computer was certainly in deprecation phase, but still has been supported.
Personally I can live comfortably without any fix or workaround for this because of at the time on the incident, I have removed the failing part of LibreOffice. But anyway, 32bit Fedora is still supported, or am I wrong? :-)
Brgds,
Eduard
This issue need to be reopened for Fedora release 30 (Thirty) While there is a workaround, this is a serious bug affecting multiple applications that depend on Java. To answer Samuel's questions: With the high price of memory, the x86 architecture is a perfect match for our lightweight VMs that only require 512Mb of RAM. Our internal tests show that it saves approx 30% over x86_64. This is due to fact that pointers are 2x as big and mostly zeros. So until x32 ABI we will remain on this architecture. This is the last I will speak of this as this conversation is going way off-topic and does not belong on a bug tracker. Last tested on kernel: 5.1.5-300.fc30.i686 *********** MASS BUG UPDATE ************** This bug is being closed with INSUFFICIENT_DATA as there has not been a response in 3 weeks. If you are still experiencing this issue, please reopen and attach the relevant data from the latest kernel you are running and any data that might have been requested previously. The needinfo request[s] on this closed bug have been removed as they have been unresolved for 1000 days |
Created attachment 1295176 [details] Back trace of libreoffice Description of problem: When launch libreoffice writer it's simply crashing. There is few ways to reproduce crash but all of them related to interacting with wiki plugin. For example it's enough to launch libreoffice and select wiki plugin in options then libreoffice crashing even without launching writer. Version-Release number of selected component (if applicable): ------------------ $ cat /etc/fedora-release Fedora release 24 (Twenty Four) ------------------ $ uname -a Linux happyfedora 4.11.8-100.fc24.i686+PAE #1 SMP Thu Jun 29 18:32:30 UTC 2017 i686 i686 i386 GNU/Linux ------------------ # dnf list libreoffice Last metadata expiration check: 2:54:28 ago on Fri Jul 7 05:12:26 2017. Installed Packages libreoffice.i686 1:5.1.6.2-8.fc24 @updates ------------------ # dnf list libreoffice-wiki-publisher Last metadata expiration check: 2:55:50 ago on Fri Jul 7 05:12:26 2017. Available Packages libreoffice-wiki-publisher.i686 1:5.1.6.2-8.fc24 updates ------------------ [root@happyfedora happiness]# dnf list java* Last metadata expiration check: 0:26:14 ago on Fri Jul 7 08:12:43 2017. Installed Packages java-1.8.0-openjdk.i686 1:1.8.0.131-3.b12.fc24 @updates-testing ------------------ How reproducible: Reproduced every time when startup writer. Or simply launch libreoffice and go to options find wiki plugin and select it. Steps to Reproduce: 1. install wiki plugin # dnf install libreoffice-wiki-publisher 2. Execute (add --backtrace to create gdbtrace.log) $ libreoffice --writer Actual results: Crash of libreoffice backtrace file attached Expected results: no crash Additional info: