1 files changed, 28 insertions, 0 deletions
diff --git a/results/classifier/zero-shot/108/performance/1784 b/results/classifier/zero-shot/108/performance/1784
new file mode 100644
index 00000000..587cea97
--- /dev/null
+++ b/results/classifier/zero-shot/108/performance/1784
@@ -0,0 +1,28 @@
+performance: 0.955
+device: 0.921
+graphic: 0.911
+other: 0.895
+permissions: 0.808
+files: 0.771
+semantic: 0.757
+network: 0.695
+boot: 0.679
+debug: 0.637
+vnc: 0.633
+socket: 0.606
+PID: 0.577
+KVM: 0.206
+
+Mac M1 Max / Debian guest / Luks password / Switching to graphical login manager (lightdm/Gdm) hangs in 75%
+Description of problem:
+In approximately 70% of cases I start QEMU with a Debian guest where the Debian guest was installed with full disk encryption, QEMU 'hangs' (does not respond') after I unlock the encrypted guest and the guest tries to start the graphical login manager (gdm or lightdm).
+
+I need to force quit QEMU, restart it multiple times until the start of the graphical login manager works.
+Steps to reproduce:
+1. Install Debian with (guided) full disk encryption and either the Gnome or the XFCE desktop environment
+2. To be able to unlock the hard disk after the installation finished, the Linux boot parameter 'console=tty1' needs to be added within grub to the Linux command line
+3. Try to restart/reboot QEMU  several times and QEMU will become unresponsive multiple times in this process.
+Additional information:
+I encounter this problem for several months now, with different versions of QEMU, macOS and Debian.
+
+There is one observation, which might help: I installed [DropBear](https://packages.debian.org/buster/dropbear-initramfs) to experiment with remote unlocking of Luks encrypted Linux boxes. It seems, that QEMU does not go into the unresponsive state, when I unlock the hard disk via SSH and not focus the QEMU window until after the graphical login manager started. (Only tried remote unlocking a few times so it is too early to confirm if this works 100% of the time.