You are not logged in.
Pages: 1
I daily use a laptop provided by my employer running Windows 7 64bit. It's a decent machine, with 8gb of RAM, an i5 processor, and an AMD Radeon HD 7570M graphics card (1gb memory).
About a year ago I set up Arch in VirtualBox, and it seemed to work pretty well. I have since had the machine re-imaged, and would like to start using Arch in VirtualBox regularly again. I'm using VirtualBox 5 now, and performance is very poor. Scrolling in Firefox, specifically, is unbearable, but all of the UI animations seem to be moving at about 1-2 fps, with substantial lag. Is anyone else having trouble with Arch Linux running as a guest the most recent version of VirtualBox?
Offline
I have what is probably the same problem, because hardware acceleration is not working correctly in Arch and VirtualBox at present.
glxinfo | grep OpenGLwill produce output including
OpenGL vendor string: VMWare, Inc
OpenGL renderer string: Gallium 0.4 on llvmpipe (LLVM 3.6, 256 bits)
OpenGL version string: 3.0 Mesa 10.6.6
OpenGL shading language version string: 1.30On my Debian VirtualBox, I get
OpenGL vendor string: Humper
OpenGL renderer string: Chromium
OpenGL version string: 2.1 Chromium 1.9
OpenGL shading language version string: 4.30 - Build 10.18.15.4256What is your desktop environment? Cinnamon warns me that I am using software rendering.
Last edited by willrogers (2015-09-15 20:42:23)
Offline
I have two virtual machines which are linked clones with slightly different setups.
One uses .xinitrc & ratpoison wm and glxinfo reports "Gallium 0.4 on llvmpipe (LLVM 3.6, 256 bits)" (software rendering).
The other uses Xfce and glxinfo reports the renderer as "Chromium" (vboxvideo driver accelerated rendering).
For the life of me I can't figure out how to get the ratpoison environment to use the Chromium renderer, but willrogers, I think this shows that acceleration is not broken entirely. We just don't know how it's configured.
Offline
are the packages for virtualbox-guest-utils & virtualbox-guest-modules installed and configured correctly ?
Read the whole VirtualBox page in archwiki, then focus on the steps outlined in chapter 2.
Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.
clean chroot building not flexible enough ?
Try clean chroot manager by graysky
Offline
Ah, sorry, I didn't really justify my statements. Yes, the packages are installed and configured correctly. Here are pastes of output for the following code:
> cat test.sh
#!/bin/bash
systemctl status vboxservice
lsmod | grep vbox
ps aux | grep -i vboxclient
glxinfo | grep 'render\|OpenGL'xfce http://pastebin.com/raw.php?i=PFiLDgp3 (vboxclient programs are duplicated because they were run as daemons, w/o --daemon flag)
ratpoison http://pastebin.com/raw.php?i=nwsh4ARA
they both report "direct rendering: Yes", but the renderer for ratpoison is the Gallium software pipe while the one for xfce is the hardware accellerated (albeit, low glsl version) Chromium renderer
lone_wolf, do i seem to have missed anything in that diagnostic script?
Offline
Dustmote,
Don't think those outputs don't gives clues about what is wrong.
Virtualbox HW acceleration appears to be enabled from outside the VM, using virtualbox settings menus .
You did mention these VMs are linked clones, please verify HW acceleration is enabled for the ratpoison VM .
If it is enabled, please post lspci -k , dmesg or journalctl -b & Xorg.0.log from both VMs.
Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.
clean chroot building not flexible enough ?
Try clean chroot manager by graysky
Offline
both have 3D accel enabled in virtualbox .. http://imgur.com/a/gVHXV (top one "browsers" is running ratpoison; bottom one "xfce" is .. xfce)
rebooting now to get shorter boot logs (i'll also verify the renderers haven't changed)
Offline
ratpoison vm
- lspci http://ix.io/mcN
- journalctl http://ix.io/mcO
- xorg http://ix.io/mcP
xfce vm
- lspci http://ix.io/mcN (same url; i think that means it got de-duplicated by the pastebin .. i've verified the original files are identical)
- journalctl http://ix.io/mcQ
- xorg http://ix.io/mcR
hmm... looking through those, it seems the machine configurations have diverged a bit... first machine uses both cpus, has fewer "no.." kernel options, etc..
Last edited by dustmote (2015-11-13 21:58:46)
Offline
Pages: 1