Summary: | Plasmashell crash in QQuickItemPrivate::setEffectiveVisibleRecur after screen change | ||
---|---|---|---|
Product: | [Plasma] plasmashell | Reporter: | NIk <nik.kaiser87> |
Component: | Desktop Containment | Assignee: | Plasma Bugs List <plasma-bugs> |
Status: | RESOLVED UPSTREAM | ||
Severity: | crash | CC: | 181500944, aggpapazach, aleixpol, alex.matte, alexander.reimelt, arpan.is, beter.peter, bharadwaj.raju777, bugseforuns, chainpunch, cjaushe4ka, fredtt695, glifwentry, im_dracula, jvgomezj, kde, lauriej, maiphi.public, marton.tamas, mediacenterkodi84, megatoaster94, mihnerts, mtjhrj, nate, nico9824, notmart, philippe, plasma-bugs, quantummhl, qydwhotmail, rudinsky, ryan9590, sitter, tampadenizen, tarcisio, tarun.aditya, valentine.stubbs, waqar.17a, yamilj, yusufnassereng |
Priority: | VHI | Keywords: | drkonqi, regression |
Version: | 5.24.0 | ||
Target Milestone: | 1.0 | ||
Platform: | openSUSE | ||
OS: | Linux | ||
See Also: |
https://bugreports.qt.io/browse/QTBUG-64249 https://bugs.kde.org/show_bug.cgi?id=419625 https://bugs.kde.org/show_bug.cgi?id=479001 |
||
Latest Commit: | Version Fixed In: | 6.0 | |
Sentry Crash Report: | |||
Attachments: |
attachment-18589-0.html
attachment-23374-0.html attachment-5119-0.html |
Description
NIk
2022-02-10 23:48:34 UTC
#49 0x000055fb02c7761c in PrimaryOutputWatcher::primaryOutputNameChanged (_t2=..., _t1=..., this=0x55fb049ff170) at /usr/src/debug/plasma5-workspace-5.24.0-1.2.x86_64/build/shell/plasmashell_autogen/EWIEGA46WW/moc_primaryoutputwatcher.cpp:136 #50 PrimaryOutputWatcher::setPrimaryOutputName (this=this@entry=0x55fb049ff170, newOutputName=...) at /usr/src/debug/plasma5-workspace-5.24.0-1.2.x86_64/shell/primaryoutputwatcher.cpp:72 #51 0x000055fb02c778f7 in operator() (__closure=<optimized out>, __closure=<optimized out>) at /usr/src/debug/plasma5-workspace-5.24.0-1.2.x86_64/shell/primaryoutputwatcher.cpp:127 primaryoutputwatcher.cpp:127 says: setPrimaryOutputName(qGuiApp->primaryScreen()->name()); primaryoutputwatcher.cpp:72 says: Q_EMIT primaryOutputNameChanged(oldOutputName, newOutputName); primaryoutputwatcher.cpp:136 says: const auto screens = qGuiApp->screens(); So we're crashing trying to get a list of all screens? Could it be empty? (In reply to Nate Graham from comment #1) > #49 0x000055fb02c7761c in PrimaryOutputWatcher::primaryOutputNameChanged > (_t2=..., _t1=..., this=0x55fb049ff170) at > /usr/src/debug/plasma5-workspace-5.24.0-1.2.x86_64/build/shell/ > plasmashell_autogen/EWIEGA46WW/moc_primaryoutputwatcher.cpp:136 > #50 PrimaryOutputWatcher::setPrimaryOutputName > (this=this@entry=0x55fb049ff170, newOutputName=...) at > /usr/src/debug/plasma5-workspace-5.24.0-1.2.x86_64/shell/ > primaryoutputwatcher.cpp:72 > #51 0x000055fb02c778f7 in operator() (__closure=<optimized out>, > __closure=<optimized out>) at > /usr/src/debug/plasma5-workspace-5.24.0-1.2.x86_64/shell/ > primaryoutputwatcher.cpp:127 > > primaryoutputwatcher.cpp:127 says: > setPrimaryOutputName(qGuiApp->primaryScreen()->name()); > > primaryoutputwatcher.cpp:72 says: > Q_EMIT primaryOutputNameChanged(oldOutputName, newOutputName); > > primaryoutputwatcher.cpp:136 says: > const auto screens = qGuiApp->screens(); > > > So we're crashing trying to get a list of all screens? Could it be empty? Dont ask me :) All those complicated lines in my report came from the automatic reporting utility under KDE, I just sent the report and dont know what any of that means. It happened today again. (In reply to Nate Graham from comment #1) > #49 0x000055fb02c7761c in PrimaryOutputWatcher::primaryOutputNameChanged > (_t2=..., _t1=..., this=0x55fb049ff170) at > /usr/src/debug/plasma5-workspace-5.24.0-1.2.x86_64/build/shell/ > plasmashell_autogen/EWIEGA46WW/moc_primaryoutputwatcher.cpp:136 > #50 PrimaryOutputWatcher::setPrimaryOutputName > (this=this@entry=0x55fb049ff170, newOutputName=...) at > /usr/src/debug/plasma5-workspace-5.24.0-1.2.x86_64/shell/ > primaryoutputwatcher.cpp:72 > #51 0x000055fb02c778f7 in operator() (__closure=<optimized out>, > __closure=<optimized out>) at > /usr/src/debug/plasma5-workspace-5.24.0-1.2.x86_64/shell/ > primaryoutputwatcher.cpp:127 > > primaryoutputwatcher.cpp:127 says: > setPrimaryOutputName(qGuiApp->primaryScreen()->name()); > > primaryoutputwatcher.cpp:72 says: > Q_EMIT primaryOutputNameChanged(oldOutputName, newOutputName); > > primaryoutputwatcher.cpp:136 says: > const auto screens = qGuiApp->screens(); > > > So we're crashing trying to get a list of all screens? Could it be empty? Edit: This (https://bugs.kde.org/show_bug.cgi?id=450062) is actually related. Just read the summary part. Its reproducible. When I do this ...: (Quote) - Main monitor connected to iGPU via HDMI, same monitor connected to GTX 1660S via DVI (GTX is bound to vfio-pci but shouldnt matter for this) - When starting a VM, the "2nd monitor" aka GTX @ DVI gets effectively disconnected from the host. When coming back to Plasma after shutting down the VM, all my desktop icons (default folder view) get reset the the top of the screen. Happens 100% of the time and only under wayland .. Plasmashell crashes. Minus the part that the icons get reset, because that happens only under Wayland, as stated under Bug 450062, not X. *** Bug 450113 has been marked as a duplicate of this bug. *** *** Bug 450607 has been marked as a duplicate of this bug. *** Suspicious commits: https://invent.kde.org/plasma/plasma-workspace/-/commit/9c04d5623687a833ed3b906451ce205b8723cedd https://invent.kde.org/plasma/plasma-workspace/-/commit/7dd5210c8766f4226c27afe5a4ef83ece6d15143 (In reply to Nate Graham from comment #1) > So we're crashing trying to get a list of all screens? Could it be empty? Very likely in my case, where the crash always happens (i.e. really every time) when the screen (there's only one) wakes up from stand-by. Syslog shows me stuff like that: A bunch of messages from kscreen saying that the screen has been disconnected, e.g. > kscreen_backend_launcher[19076]: kscreen.xcb.helper: Connection: "Disconnected" Several of these, 6x with screen -1 and a lot with screen 0: > plasmashell[929]: requesting unexisting screen -1 > plasmashell[929]: requesting unexisting screen 0 Then again kscreen reports about the disconnected screen and within a second notices that it has been connected again. After that plasmashell restarts. After updating to KDE 5.24.2 on openSUSE Tumbleweed snapshot 20220227 I would carefully say that the issue has been resolved. I tested multiple times to start my VMs, restarted 3 times, tried the same again, plasmashell did not crash this time. I just wonder why there is no "solved" or merge request thingy in here, or does that appear later? The other duplicates are also still open. You can mark the bug as "RESOLVED WORKSFORME". (In reply to Fushan Wen from comment #10) > You can mark the bug as "RESOLVED WORKSFORME". Thanks The "duplicate" bug filed in https://bugs.kde.org/show_bug.cgi?id=450607 has not been fixed in Tumbleweed 20220227. I'm not saying that "WORKSFORSOME" is not correct in the sense that one possible trigger of the underlying bug may have been fixed. But I fear that setting the status to "WORKSFORSOME" reduces the priority even further. Given the fact that a core component of the DE is frequently crashing in everyday usage this might send the wrong signal. Created attachment 147244 [details] attachment-18589-0.html I must chime in and say that the frequent crashing for various unpredictable reasons not just this particular bug, that was more easily reproducible, has force me to abandon KDE and Plasma. I am using OpenBox and XFCE and as yet have not had this DM crash once. If this is obtainable in this environment, I fear there is and underlying issue yet to be tracked down, and setting the status to something that will place the issue further on the back burner will only mean the core trigger is never found let alone be resolved. On 3/2/22 06:54, Philipp Maierhöfer wrote: > https://bugs.kde.org/show_bug.cgi?id=449981 > > --- Comment #12 from Philipp Maierhöfer<maiphi.public@gmx.net> --- > The "duplicate" bug filed inhttps://bugs.kde.org/show_bug.cgi?id=450607 has > not been fixed in Tumbleweed 20220227. > > I'm not saying that "WORKSFORSOME" is not correct in the sense that one > possible trigger of the underlying bug may have been fixed. But I fear that > setting the status to "WORKSFORSOME" reduces the priority even further. > > Given the fact that a core component of the DE is frequently crashing in > everyday usage this might send the wrong signal. > > frequent crashing for various unpredictable reasons not just this particular bug
Can you file different bug reports about the other issues?
(In reply to Philipp Maierhöfer from comment #12) > The "duplicate" bug filed in https://bugs.kde.org/show_bug.cgi?id=450607 has > not been fixed in Tumbleweed 20220227. > > I'm not saying that "WORKSFORSOME" is not correct in the sense that one > possible trigger of the underlying bug may have been fixed. But I fear that > setting the status to "WORKSFORSOME" reduces the priority even further. > > Given the fact that a core component of the DE is frequently crashing in > everyday usage this might send the wrong signal. I agree, changed back to reported if thats correct. I assumed that since its actually fixed for me, consistently, it also is fixed for anyone else but they just didnt report back. The plasmashell crash when the screen wakes up from energy saving mode doesn't seem to happen anymore in 5.24.3. Created attachment 147585 [details] attachment-23374-0.html you might as well put me down as satisfied. I have long moved on from using plasma5 / KDE. Way many issues and way too much of a resource hog. Why spend 6G running just your shell when you can accomplish exactly the same thing and only spend 2.5-3G. Doesn't matter on my machines, but what about someone you has less than 32G or 16G, what about a machine that only has 8G for example On 3/18/22 11:39, Philipp Maierhöfer wrote: > https://bugs.kde.org/show_bug.cgi?id=449981 > > --- Comment #16 from Philipp Maierhöfer<maiphi.public@gmx.net> --- > The plasmashell crash when the screen wakes up from energy saving mode doesn't > seem to happen anymore in 5.24.3. > The crash when monitors wake up from sleep still happens on two monitors connected via DP. Operating System: Arch Linux KDE Plasma Version: 5.24.3 KDE Frameworks Version: 5.92.0 Qt Version: 5.15.3 Kernel Version: 5.16.16-zen1-1-zen (64-bit) Graphics Platform: Wayland Processors: 12 × AMD Ryzen 5 2600X Six-Core Processor Memory: 15.6 GiB of RAM Graphics Processor: AMD Radeon RX 580 Series Are you able to reproduce this issue in Plasma 5.25 or later? Created attachment 151932 [details] attachment-5119-0.html I no longer use KDE or Plasma. For that matter I no longer use openSUSE. I have switch to a much more stable Fedora currently version 36, and use XFCE as a DM, and my base memory usage is 3.18 GiB with Firefox and Thunderbird loaded as apposed to over 8 GiB with Plasma 5 On 2022-09-08 14:13, Nate Graham wrote: > https://bugs.kde.org/show_bug.cgi?id=449981 > > Nate Graham<nate@kde.org> changed: > > What |Removed |Added > ---------------------------------------------------------------------------- > Status|REPORTED |NEEDSINFO > Resolution|--- |WAITINGFORINFO > > --- Comment #19 from Nate Graham<nate@kde.org> --- > Are you able to reproduce this issue in Plasma 5.25 or later? > Dear Bug Submitter, This bug has been in NEEDSINFO status with no change for at least 15 days. Please provide the requested information as soon as possible and set the bug status as REPORTED. Due to regular bug tracker maintenance, if the bug is still in NEEDSINFO status with no change in 30 days the bug will be closed as RESOLVED > WORKSFORME due to lack of needed information. For more information about our bug triaging procedures please read the wiki located here: https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging If you have already provided the requested information, please mark the bug as REPORTED so that the KDE team knows that the bug is ready to be confirmed. Thank you for helping us make KDE software even better for everyone! *** Bug 459553 has been marked as a duplicate of this bug. *** *** Bug 450594 has been marked as a duplicate of this bug. *** *** Bug 450723 has been marked as a duplicate of this bug. *** Is anyone else affected by this able to reproduce it in Plasma 5.25? I notice all reports are from 5.24. Dear Bug Submitter, This bug has been in NEEDSINFO status with no change for at least 15 days. Please provide the requested information as soon as possible and set the bug status as REPORTED. Due to regular bug tracker maintenance, if the bug is still in NEEDSINFO status with no change in 30 days the bug will be closed as RESOLVED > WORKSFORME due to lack of needed information. For more information about our bug triaging procedures please read the wiki located here: https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging If you have already provided the requested information, please mark the bug as REPORTED so that the KDE team knows that the bug is ready to be confirmed. Thank you for helping us make KDE software even better for everyone! This bug has been in NEEDSINFO status with no change for at least 30 days. The bug is now closed as RESOLVED > WORKSFORME due to lack of needed information. For more information about our bug triaging procedures please read the wiki located here: https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging Thank you for helping us make KDE software even better for everyone! *** Bug 400835 has been marked as a duplicate of this bug. *** *** Bug 402121 has been marked as a duplicate of this bug. *** *** Bug 404902 has been marked as a duplicate of this bug. *** *** Bug 410575 has been marked as a duplicate of this bug. *** *** Bug 417232 has been marked as a duplicate of this bug. *** *** Bug 428573 has been marked as a duplicate of this bug. *** *** Bug 434823 has been marked as a duplicate of this bug. *** *** Bug 440265 has been marked as a duplicate of this bug. *** *** Bug 441728 has been marked as a duplicate of this bug. *** *** Bug 459958 has been marked as a duplicate of this bug. *** *** Bug 468952 has been marked as a duplicate of this bug. *** Got a report from a KDE dev using Plasma 5.27; re-opening. The latest qt5-declarative has reverted the backport commit from Qt6, so it's expected the bug is not fixed in Qt5. This one? https://invent.kde.org/qt/qt/qtdeclarative/-/commit/2f14549ee64782e9d64257696b10c6d59e884a7d I believe so, yeah. Plasma 6 is in progress, so I think at this point we'll have to be okay with it being a Plasma-6-only fix. *** Bug 473202 has been marked as a duplicate of this bug. *** *** Bug 471057 has been marked as a duplicate of this bug. *** *** Bug 471220 has been marked as a duplicate of this bug. *** *** Bug 443901 has been marked as a duplicate of this bug. *** *** Bug 445921 has been marked as a duplicate of this bug. *** *** Bug 462431 has been marked as a duplicate of this bug. *** *** Bug 472521 has been marked as a duplicate of this bug. *** *** Bug 474180 has been marked as a duplicate of this bug. *** *** Bug 481662 has been marked as a duplicate of this bug. *** *** Bug 489659 has been marked as a duplicate of this bug. *** *** Bug 490964 has been marked as a duplicate of this bug. *** *** Bug 492169 has been marked as a duplicate of this bug. *** *** Bug 481901 has been marked as a duplicate of this bug. *** *** Bug 491874 has been marked as a duplicate of this bug. *** *** Bug 487510 has been marked as a duplicate of this bug. *** *** Bug 495662 has been marked as a duplicate of this bug. *** *** Bug 496559 has been marked as a duplicate of this bug. *** |