Created attachment 121364 [details] strace of a slow startup SUMMARY STEPS TO REPRODUCE 1. start dophin 2. what until it shows up OBSERVED RESULT EXPECTED RESULT SOFTWARE/OS VERSIONS Windows: macOS: Linux/KDE Plasma: (available in About System) KDE Plasma Version: KDE Frameworks Version: Qt Version: ADDITIONAL INFORMATION
there are thousands of lines of 4498 13:05:35.209272 read(10, "\2\0\0\0\2\0\0\0\0\0\0\0\20\0\0\0log1.txt\0\0\0\0\0\0\0\0", 32) = 32 <0.000008> 14498 13:05:35.209317 poll([{fd=5, events=POLLIN}, {fd=6, events=POLLIN}, {fd=10, events=POLLIN}, {fd=11, events=POLLIN}, {fd=12, events=POLLPRI}, {fd=13, events=POLLIN}, {fd=19, events=POLLIN}, {fd=20, events=POLLIN}, {fd=21, events=POLLIN}, {fd=22, events=POLLIN}], 10, 104) = 1 ([{fd=10, revents=POLLIN}]) <0.000009> 14498 13:05:35.209395 ioctl(10, FIONREAD, [32]) = 0 <0.000007> log1.txt was the file strace stored its output into, and dolphin seems to poll this files
According to the strace log, a futex with a timeout value of 5 seconds was aborted with ETIMEDOUT. This might help to investigate the deadlock.
There are two issues causing the bug, I believe : What Christopher said, we can see the 5 seconds futex timeout : 14498 13:05:27.346112 futex(0x55638e72d680, FUTEX_WAIT_PRIVATE, 0, {tv_sec=4, tv_nsec=999998802} <unfinished ...> 14499 13:05:27.346137 poll([{fd=3, events=POLLIN}], 1, -1 <unfinished ...> 14498 13:05:32.346393 <... futex resumed> ) = -1 ETIMEDOUT (Connection timed out) <5.000231> 14498 13:05:32.346666 futex(0x55638e72d630, FUTEX_WAKE_PRIVATE, 1) = 0 <0.000097> The futex lock seems to be in ki18n loading, because prior to this mutex loop invocation we have : 14498 13:05:27.341563 statx(AT_FDCWD, "/usr/share/locale/en/LC_SCRIPTS/kio5/kio5.js", AT_STATX_SYNC_AS_STAT, STATX_ALL, 0x7ffcb000c4d0) = -1 ENOENT (No such file or directory) <0.000012> That is clearly originating from klocalizedstring.cpp and this file deels with QMutex. The lines : 4498 13:05:35.209272 read(10, "\2\0\0\0\2\0\0\0\0\0\0\0\20\0\0\0log1.txt\0\0\0\0\0\0\0\0", 32) = 32 <0.000008> 14498 13:05:35.209317 poll([{fd=5, events=POLLIN}, {fd=6, events=POLLIN}, {fd=10, events=POLLIN}, {fd=11, events=POLLIN}, {fd=12, events=POLLPRI}, {fd=13, events=POLLIN}, {fd=19, events=POLLIN}, {fd=20, events=POLLIN}, {fd=21, events=POLLIN}, {fd=22, events=POLLIN}], 10, 104) = 1 ([{fd=10, revents=POLLIN}]) <0.000009> 14498 13:05:35.209395 ioctl(10, FIONREAD, [32]) = 0 <0.000007> Could be due to mimetype determination or metadata gathering going wrong, reading a whole file when it fails or errors. I haven't yet investigated into it as much as the first issue. I reproduce at least one of the two issues, but I don't know which one yet.
dupe of bug 348521 ?
(In reply to Patrick Silva from comment #4) > dupe of bug 348521 ? No, we need to keep bugs separate when the deal with different things even similar. Having meta-bugs non-discerning bugs where no one has the same symptoms make it so very hard to fix. sometimes takes 1 min / 30ses != starting is a bit slow
*** Bug 416937 has been marked as a duplicate of this bug. ***
fixed some time ago