summary refs log tree commit diff stats
path: root/results/classifier/accel-gemma3:12b/vmm/1750899
diff options
context:
space:
mode:
Diffstat (limited to 'results/classifier/accel-gemma3:12b/vmm/1750899')
-rw-r--r--results/classifier/accel-gemma3:12b/vmm/175089925
1 files changed, 25 insertions, 0 deletions
diff --git a/results/classifier/accel-gemma3:12b/vmm/1750899 b/results/classifier/accel-gemma3:12b/vmm/1750899
new file mode 100644
index 000000000..1ca30c3f8
--- /dev/null
+++ b/results/classifier/accel-gemma3:12b/vmm/1750899
@@ -0,0 +1,25 @@
+
+Mouse cursor sometimes can't pass the invisible border on the right side of the screen
+
+I'm using qemu 2.11 on Gentoo Linux, with configured GPU passthrough (Radeon RX580) to the guest Windows 10.
+This configuration is alive for last 4 years, this time I changed a lot qemu, linux kernel and windows versions, changed GPU and always all was working as expected. I always used standard PS/2 mouse emulation and that was enough for me.
+
+Now, I bought two new monitors, instead of old one, and setup them as one logical monitor, using technology called Eyefinity - it's a part of standard Radeon software. Now Windows thinks, that I have one monitor with resolution 2160x1920 (I bought Dell monitors with a thin borders and use them in portrait mode).
+
+Windows uses it without any problems, but mouse become crazy - sometimes (~3 times from each 5) I can't move cursor to the right border of the screen, it looks like the invisible vertical border. I spent really huge amount of time to understand, which component is the root of problem and found, that it's really a mouse. I tried all possible variants (standard, tablet, virtio-mouse-pci, virtio-tablet-pci), and found, that in both mouse variants bug is reproducing, and in both tablet variants - cursor stuck near all real borders and corners, so it's not a variant too.
+The only working variant becomes passing real USB port to my VM and insert second mouse to this port. So, now it's working, but I have two mice on my working place, which doesn't seems very useful.
+
+Here is my command line:
+
+QEMU_AUDIO_DRV=pa QEMU_PA_SAMPLES=4096 qemu-system-x86_64 -enable-kvm -M q35 -m 12168 -cpu host,kvm=off -smp 4,sockets=1,cores=4 \
+-bios /usr/share/qemu/bios.bin -rtc base=localtime -vga none -device secondary-vga \
+-drive id=virtiocd,if=none,format=raw,file=/home/akushsky/virtio-win-0.1.141.iso \
+-device driver=ide-cd,bus=ide.1,drive=virtiocd \
+-device ioh3420,bus=pcie.0,addr=1c.0,multifunction=on,port=1,chassis=1,id=root.1 \
+-device vfio-pci,host=05:00.0,bus=root.1,addr=00.0,multifunction=on,romfile=/opt/kvm/images/Sapphire.RX580.8192.170320_1.bin,x-vga=on \
+-device virtio-scsi-pci,id=scsi \
+-drive file=/dev/sdb,id=disk,format=raw,if=none,discard=on,cache=none,aio=native,detect-zeroes=unmap -device scsi-hd,drive=disk,id=scsi0 \
+-device ich9-intel-hda,bus=pcie.0,addr=1b.0,id=sound0 -device hda-duplex,id=sound0-codec0,bus=sound0.0,cad=0 \
+-usb -usbdevice host:046d:c52b
+
+All in all, I checked on Windows 7 and Windows 10, and on qemu 2.10 and 2.11 - bug is always reproducible.
\ No newline at end of file