Note: This is a public test instance of Red Hat Bugzilla. The data contained within is a snapshot of the live data so any changes you make will not be reflected in the production Bugzilla. Email is disabled so feel free to test any aspect of the site that you want. File any problems you find or give feedback at bugzilla.redhat.com.
Bug 1206960
Summary: | Various apps crash with an X BadMatch error when run on GNOME with llvmpipe | ||||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | Wallace Hermano <wallacehermano> | ||||||||||||||||||||||||
Component: | clutter | Assignee: | Igor Gnatenko <ignatenko> | ||||||||||||||||||||||||
Status: | CLOSED ERRATA | QA Contact: | Fedora Extras Quality Assurance <extras-qa> | ||||||||||||||||||||||||
Severity: | unspecified | Docs Contact: | |||||||||||||||||||||||||
Priority: | unspecified | ||||||||||||||||||||||||||
Version: | 22 | CC: | airlied, ajax, awilliam, ccecchi, danofsatx, fmuellner, fschwarz, ignatenko, itamar, jonha87, kalevlember, marcelo.barbosa, mcatanzaro+wrong-account-do-not-cc, mclasen, msrb, otaylor, pbrobinson, r, robatino, sergio, tobias47n9e, wallacehermano, walters | ||||||||||||||||||||||||
Target Milestone: | --- | ||||||||||||||||||||||||||
Target Release: | --- | ||||||||||||||||||||||||||
Hardware: | x86_64 | ||||||||||||||||||||||||||
OS: | Unspecified | ||||||||||||||||||||||||||
URL: | https://retrace.fedoraproject.org/faf/reports/bthash/ba6a1988970018f2de2a50d74575517d7a8cb4f9 | ||||||||||||||||||||||||||
Whiteboard: | AcceptedBlocker abrt_hash:6b6792aeb7e670b73f3fff6a12f98bf7a0903d24 | ||||||||||||||||||||||||||
Fixed In Version: | clutter-1.22.0-2.fc22 | Doc Type: | Bug Fix | ||||||||||||||||||||||||
Doc Text: | Story Points: | --- | |||||||||||||||||||||||||
Clone Of: | Environment: | ||||||||||||||||||||||||||
Last Closed: | 2015-05-11 19:06:35 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: | |||||||||||||||||||||||||||
Bug Blocks: | 1043130 | ||||||||||||||||||||||||||
Attachments: |
|
Description
Wallace Hermano
2015-03-30 00:27:30 UTC
Created attachment 1008107 [details]
File: backtrace
Created attachment 1008108 [details]
File: cgroup
Created attachment 1008109 [details]
File: core_backtrace
Created attachment 1008110 [details]
File: dso_list
Created attachment 1008111 [details]
File: environ
Created attachment 1008112 [details]
File: limits
Created attachment 1008113 [details]
File: maps
Created attachment 1008114 [details]
File: mountinfo
Created attachment 1008115 [details]
File: namespaces
Created attachment 1008116 [details]
File: open_fds
Created attachment 1008117 [details]
File: proc_pid_status
Another user experienced a similar problem: Can be reproduced reliably in a relatively fresh Fedora 22 installation. Happens everytime I start the "Maps" activity. Not running a Wayland session. reporter: libreport-2.5.1 backtrace_rating: 4 cmdline: /usr/bin/gjs /usr/share/gnome-maps/org.gnome.Maps --gapplication-service crash_function: _g_log_abort executable: /usr/bin/gjs-console global_pid: 32003 kernel: 4.0.0-0.rc5.git4.1.fc22.x86_64 package: gjs-1.43.3-1.fc22 reason: gjs-console killed by SIGTRAP runlevel: N 5 type: CCpp uid: 1000 Another user experienced a similar problem: Occurs on launch of Maps from a clean install of Fedora 22 Final TC1 Workstation live x86_64. reporter: libreport-2.5.1 backtrace_rating: 4 cmdline: /usr/bin/gjs /usr/share/gnome-maps/org.gnome.Maps --gapplication-service crash_function: _g_log_abort executable: /usr/bin/gjs-console global_pid: 5597 kernel: 4.0.0-1.fc22.x86_64 package: gjs-1.43.3-1.fc22 reason: gjs-console killed by SIGTRAP runlevel: N 5 type: CCpp uid: 1000 Proposed Final blocker: "All applications that can be launched using the standard graphical mechanism of a release-blocking desktop after a default installation of that desktop must start successfully and withstand a basic functionality test." - https://fedoraproject.org/wiki/Fedora_22_Final_Release_Criteria#Default_application_functionality OK, so this and the recent cases of https://bugzilla.redhat.com/show_bug.cgi?id=1200559 are all likely the same bug, and the bug is this: when you run the app with llvmpipe, it crashes with an X BadMatch error. You can reproduce this by running them in a KVM, on a metal system with no hardware acceleration, or on a system *with* acceleration like this: LIBGL_ALWAYS_SOFTWARE=1 totem LIBGL_ALWAYS_SOFTWARE=1 gnome-maps I've confirmed both totem and Maps are affected on a clean boot of Final TC2 Workstation x86_64 live - no need to install, it can be reproduced right from the live environment. https://dl.fedoraproject.org/pub/alt/stage/22_TC2/Workstation/x86_64/iso/Fedora-Live-Workstation-x86_64-22-TC2.iso drago01 said file against mesa, so moving to that. Reported in upstream. https://bugs.freedesktop.org/show_bug.cgi?id=90354 I think we can go ahead and mark this as accepted blocker, since bug #1200559 was previously accepted blocker. Moving the blocker status from that bug to this one is just Bugzilla wrangling and shouldn't require a formal meeting. Actually I'd rather re-discuss it, because the situation has changed a bit. That one was approved on the belief that it *always happened*, but in fact it only happens when hardware acceleration isn't available, so it's not quite the same situation we discussed. Also affects Cheese, I believe. Looks like some disagreement between a 32-bit depth and 24-bit depth somewhere, so this might not be a mesa bug, bug a cogl/clutter bug. Basically swrast is trying to putimage a 24 depth/ 32 bpp image to a 32/32 drawable. I've debugged this a big, pretty sure its a cogl/clutter/gtk bug, a 32-bit window gets created, then cogl creates a 24-bit glX #0 0x000000302bc410d0 in XCreateWindow (dpy=dpy@entry=0x87fc00, parent=parent@entry=31457314, x=-1, y=46, width=1, height=1, borderWidth=0, depth=32, class=1, visual=0x87f3c0, valuemask=8218, attributes=0x7fffffff7da0) at Window.c:113 #1 0x0000003d3b8698e5 in _gdk_x11_display_create_window_impl (display=0x890000 [GdkX11Display], window=0xf55e70 [GdkX11Window], real_parent=<optimized out>, screen=0x896020 [GdkX11Screen], event_mask=14827270, attributes=0x0, attributes_mask=0) at gdkwindow-x11.c:1101 #2 0x0000003d3b842d3d in gdk_window_ensure_native (window=0xf55e70 [GdkX11Window]) at gdkwindow.c:1756 #3 0x0000003d3f244130 in clutter_stage_gdk_realize (stage_window=0xc2d190) at gdk/clutter-stage-gdk.c:267 #4 0x0000003d3f2ae58c in clutter_stage_realize (self=0xc63210 [ClutterStage]) at clutter-stage.c:746 #5 0x000000302a40fe74 in _g_closure_invoke_va (closure=closure@entry=0x612440, return_value=return_value@entry=0x0, instance=instance@entry=0xc63210, args=args@entry=0x7fffffff8180, n_params=<optimized out>, param_types=0x0) at gclosure.c:831 XID 0x1e000ec #0 0x00007ffff7b25279 in XCreateDrawable (pdp=0xcfa3e0, dpy=0x87fc00, drawable=31457516, visualid=537) at drisw_glx.c:60 #1 0x00007ffff7b25f09 in driswCreateDrawable (base=0x913b10, xDrawable=31457516, drawable=31457517, modes=0x9e8200) at drisw_glx.c:516 #2 0x00007ffff7af458a in CreateDRIDrawable (dpy=0x87fc00, config=0x9e8200, drawable=31457516, glxdrawable=31457517, attrib_list=0x0, num_attribs=0) at glx_pbuffer.c:209 #3 0x00007ffff7af4e34 in CreateDrawable (dpy=0x87fc00, config=0x9e8200, drawable=31457516, attrib_list=0x0, glxCode=31 '\037') at glx_pbuffer.c:492 #4 0x00007ffff7af5759 in glXCreateWindow (dpy=0x87fc00, config=0x9e8200, win=31457516, attrib_list=0x0) at glx_pbuffer.c:967 #5 0x0000003d4266d752 in _cogl_winsys_onscreen_init (onscreen=0xfc55d0, error=0x7fffffff7e90) at winsys/cogl-winsys-glx.c:1421 #6 0x0000003d4265a8de in cogl_framebuffer_allocate (framebuffer=0xfc55d0, error=error@entry=0x7fffffff7e90) at ./cogl-framebuffer.c:750 cogl asks for visual 537 which is 24-bit. So not mesa bug from what I can see. reassign to gtk for now Upstream clutter fix here: https://git.gnome.org/browse/clutter/commit/?id=8fc8a73580ceb810ee7f34eae0df26fac4cb277d clutter-1.22.0-2.fc22 has been submitted as an update for Fedora 22. https://admin.fedoraproject.org/updates/clutter-1.22.0-2.fc22 Fix confirmed in TC3. clutter-1.22.0-2.fc22 has been pushed to the Fedora 22 stable repository. If problems still persist, please make note of it in this bug report. Discussed at the 2015-05-11 blocker review meeting[0]. Voted as AcceptedBlocker. #agreed 1206960 - AcceptedBlocker - violates "All applications that can be launched using the standard graphical mechanism of a release-blocking desktop after a default installation of that desktop must start successfully and withstand a basic functionality test" if hardware acceleration isn't available [0]: http://meetbot.fedoraproject.org/fedora-blocker-review/2015-05-11/ Correcting erroneous resolution change. *** Bug 1221715 has been marked as a duplicate of this bug. *** (In reply to Fedora Update System from comment #26) > clutter-1.22.0-2.fc22 has been pushed to the Fedora 22 stable repository. > If problems still persist, please make note of it in this bug report. totem and maps still crash here with clutter-1.22.2-2.fc22.x86_64 Do you need further information? totem still crash for me too, we got bug #1206027 opened since 23 March, 2015 Today was an update to clutter-gtk-1.6.2-1.fc22.x86_64 clutter-gst2-2.0.14-2.fc22.x86_64 clutter-gst3-3.0.8-1.fc22.x86_64 clutter-1.22.4-1.fc22.x86_64 xorg-x11-drv-nvidia-352.30-1.fc22.x86_64 kmod-nvidia-4.1.3-200.fc22.x86_64-352.30-2.fc22.x86_64 and now Maps and totem crash again… |