Bug 428838
Summary: | Update request: conduit-0.3.6 | ||||||
---|---|---|---|---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | Michel Lind <michel> | ||||
Component: | conduit | Assignee: | Bernard Johnson <bjohnson> | ||||
Status: | CLOSED CURRENTRELEASE | QA Contact: | Fedora Extras Quality Assurance <extras-qa> | ||||
Severity: | low | Docs Contact: | |||||
Priority: | medium | ||||||
Version: | 8 | CC: | john.carr, tim.lauridsen | ||||
Target Milestone: | --- | ||||||
Target Release: | --- | ||||||
Hardware: | All | ||||||
OS: | Linux | ||||||
Whiteboard: | |||||||
Fixed In Version: | 0.3.11.1-1.fc8 | Doc Type: | Bug Fix | ||||
Doc Text: | Story Points: | --- | |||||
Clone Of: | Environment: | ||||||
Last Closed: | 2008-07-06 06:15:25 UTC | Type: | --- | ||||
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: | 429970 | ||||||
Bug Blocks: | |||||||
Attachments: |
|
Description
Michel Lind
2008-01-15 16:02:11 UTC
Created attachment 291722 [details]
Patch to conduit spec file
Not sure this is the way to go (overriding %{_libdir} -- but otherwise if the
package is built on 64-bit machines it will install files in a different
location), so I have not patched the remaining shebang-scripts.
Would you be looking for a comaintainer, by any chance? This package looks
really interesting. Thanks.
I have a build that appears to work, but when I exit, I get: python: Modules/gcmodule.c:276: visit_decref: Assertion `gc->gc.gc_refs != 0' failed. 4a8de83b-5090-5b69-3b181b5a-3b235430 is dumped Is that something you've seen in your build? I'm reluctant to push any build, even to devel, while this is appearing. Here is my build for F-8: http://koji.fedoraproject.org/koji/taskinfo?taskID=366851 If you want to help with the package, request permissions in packagedb and I'll set you up. Mine has the assertion failure, but nothing is dumped: python: Modules/gcmodule.c:276: visit_decref: Assertion `gc->gc.gc_refs != 0' failed. It might be better to wait for 0.4.0 then? That's the next stable version. I've asked for permission in packagedb, thanks. It looks to be a python bug since there is no native code in conduit. If you have time, install the python-debuginfo files and get a stacktrace. Although this doesn't seems to be a conduit bug, I don't want to do any releases on this until we have a better understanding of the problem. It would cause a bunch of bug reports to be filed. Look like there have been released a 0.3.6 version. http://www.conduit-project.org/wiki/0.3.6 If you guys need somebody to test the package, then just let med know The build in comment #2 is 0.3.6 if you want to test drive it. I'm assuming you'll run into bug #429970 though which is currently blocking a new release. Michel, if I understand your comment on bug #429970, you have tested the build on rawhide and it does not crash? If that is so, I'm going to push out a build to rawhide to get wider testing/feedback. I have not tested that particular build -- there does not seem to be any way to download it. I'm still on my hacked-together 0.3.5, which also suffered from the assertion failures before I upgraded to Rawhide. But yes, push it and I'll give it a spin, thanks. I have made a local build for F8 of 0.3.6 located here. http://timlau.fedorapeople.org/files/test/conduit/ i get the python: Modules/gcmodule.c:276: visit_decref: Assertion `gc->gc.gc_refs != 0' failed. on exit. I look like the Flickr sync i working in 0.3.6 :) See #509702 on bgo http://bugzilla.gnome.org/show_bug.cgi?id=509702 (In reply to comment #8) > But yes, push it and I'll give it a spin, thanks. Here is a rawhide build: http://koji.fedoraproject.org/koji/taskinfo?taskID=389742 conduit-0.3.8-1.fc8 has been submitted as an update for Fedora 8 conduit-0.3.8-1.fc8 has been pushed to the Fedora 8 testing repository. If problems still persist, please make note of it in this bug report. If you want to test the update, you can install it with su -c 'yum --enablerepo=updates-testing update conduit'. You can provide feedback for this update here: http://admin.fedoraproject.org/updates/F8/FEDORA-2008-1972 In testing, I'm getting a SEGFAULT: [Main ][INFO ] Conduit v0.3.8 Installed: True (Main.py:79) [Main ][INFO ] Using UI: gtk (Main.py:80) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules (Module.py:103) [Module ][DEBUG ] Reading directory /home/rwood/.conduit/modules (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/N800Module (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/BackpackModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/FspotModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/BansheeModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/FlickrModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/ShutterflyModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/NetworkModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/RhythmboxModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/EvolutionModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/SmugMugModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/FileModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/GConfModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/FacebookModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/GmailModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/FeedModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/BoxDotNetModule (Module.py:103) [Module ][DEBUG ] Reading directory /usr/lib/python2.5/site-packages/conduit/modules/GoogleModule (Module.py:103) [dataproviders.SimpleFactory][INFO ] HAL Initialized (VolumeFactory.py:25) [dataproviders.SimpleFactory][INFO ] HAL Initialized (VolumeFactory.py:25) [Utils ][DEBUG ] Adding /usr/lib/python2.5/site-packages/conduit/modules/BackpackModule/backpack to python path (Utils.py:117) python: Modules/gcmodule.c:276: visit_decref: Assertion `gc->gc.gc_refs != 0' failed. /usr/lib/firefox-2.0.0.12/run-mozilla.sh: line 131: 6242 Segmentation fault "$prog" ${1+"$@"} Can someone confirm whether or not this is: http://bugzilla.gnome.org/show_bug.cgi?id=509702 There is a test case in http://bugzilla.gnome.org/show_bug.cgi?id=509702#c28 If that causes the crash, then the bug is in your python gnomevfs bindings. (As this has not been reported on anything but RH/Fedora and PLD Linux...) Tried the test case you referenced on a fully updated F8: [bjohnson@localhost ~]$ python Python 2.5.1 (r251:54863, Oct 30 2007, 13:54:11) [GCC 4.1.2 20070925 (Red Hat 4.1.2-33)] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> def _(*args, **kwargs): ... pass ... >>> import gnomevfs >>> t1 = gnomevfs.VolumeMonitor() >>> t1.connect("volume-mounted",_) 1L >>> t2 = gnomevfs.VolumeMonitor() >>> t2.connect("volume-mounted",_) 2L >>> Although, when I run it non-interactively: [bjohnson@localhost tmp]$ python t python: Modules/gcmodule.c:276: visit_decref: Assertion `gc->gc.gc_refs != 0' failed. Aborted [bjohnson@localhost tmp]$ So it's a bug in the gnomevfs bindings rather than conduit. Most applications don't hit this because they only have one volume monitor instance, but because of our plugins we've ended up with 2 - thus triggering the bug. The assertion hints that we are allocating something more than once, but only freeing it once, so the reference count isn't 0 with python is closing. It seems to only happen on F8/PLD Linux boxes so far, so i don't want to report it directly GNOME bugzilla... Is gnome-vfs2 the right component for this bug? I have an update to conduit that disables the iPod and N800 modules for now (enabling any of them trigger the gnome-vfs bug). Also updates the iPod dependency to python-gpod, not libgpod, and fixes the Firefox detection to be multilib-aware. conduit-0.3.11.1-1.fc9 has been submitted as an update for Fedora 9 conduit-0.3.11.1-1.fc8 has been submitted as an update for Fedora 8 conduit-0.3.11.1-1.fc8 has been pushed to the Fedora 8 stable repository. If problems still persist, please make note of it in this bug report. conduit-0.3.11.1-1.fc9 has been pushed to the Fedora 9 stable repository. If problems still persist, please make note of it in this bug report. |