https://bugs.kde.org/show_bug.cgi?id=406810
--- Comment #29 from Henning <li...@hkopf.de> --- Assuming that the Dolphin+EncFS-problem is a result of the CPU/CPU-flags in conjunction with the compiler/compiler-flags I also did some further investigations. Out of habit I used "Oracle VM Virtualbox" with host-extensions installed for USB support. (No guest-extensions installed by me; Virtualbox 6.1) The results: i5-4200U host: Kubuntu 20.04.1 works fine with Dolphin+EncFS VM on that i5-4200U host: Kubuntu 20.04.1 shows the Dolphin+EncFS-problem i7-3930K host: Kubuntu 20.04.1 shows the Dolphin+EncFS-problem Therefore I took a look at the CPU-flags by comparing the output of /proc/cpuinfo. If I can attach the LO-calc-file, I will do so. But from my point of view, these flags are the most interesting: Flag only on the two systems where the problem appears (VM@4200U and 3930K): x2apic Flags missing on the two systems where the problem appears but available on the working system (4200U): bmi1, bmi2, cpuid_fault, ept_ad, erms, f16c, fma, sdbg, smep, tsc_adjust. There are a lot flags missing only in the VM and some others are missing on the 3930K only. Hence I listed all flags that are common on the systems where the Dolphin+EncFS-problem appears. Since I don't really know much about CPU- or compiler-flags, I also played around with CPU-microcode and patches on the native 3930K. Booting without any microcode loaded to the CPU by Kubuntu does not solve the Dolphin+EncFS-problem just as adding "mitigations=off" to grub. Does anyone have an old ISO-image of Neon with Plasma 5.18? Testing Dolphin+EncFS there inside a VM would be quite interesting! P.S.: A Kubuntu 20.04.1-VM on the i5-4200U running Win10 as host OS also shows the Dolphin+EncFS-problem. -- You are receiving this mail because: You are watching all bug changes.
