gnif Thanks for all your hard work that went into making Looking Glass possible. I’ll put it into the description to let people know to use the B2 release
Thanks a ton! A note: LG B5.0.1 (and possibly earlier) needed the xpresent-dev package installed (via standard package manager) as well in order for the compilation to work.
Thanks for the video and for the script! Everything worked, as it should be on PoP OS 20.04. I always wanted to do hotswappable peripherals setup without a real physical KVM. And, here with this software, plus a monitor with two plugs, I am just happy ;D!
Hello! Thanks for the brilliant guide. It has helped me to get Looking Glass up and running! One quick note that could be helpful: I had to buy a new GPU to replace my GTX660 that doesn't support UEFI. It seems as though the Nvidia 43 error cannot be avoided when using default SeaBIOS, so you must use the UEFI one instead (hence rendering old graphics cards unusable). Cheers.
It should be possible to use Seabios with Nvidia GPUs. I have a GTX580 that I use every now and then. In my Windows XP video I did a couple extra steps to get a GTX 980 working under Seabios.
Just finished installing it on Ubuntu 20.04 and the last LG release. Work perfectly for my usage in 3440*1440 with 64M of memory. (passthrough GPU a GTX1080) I'm using it to use my elgato 60S and game capture to play with my Switch. Since i use an external audio card, audio latency is less than 10ms avec according to the calibration of games like crypt of necrodancer, visual latency is under 50ms wich is fine. Many thanks to you.
because with GPU pass through you will have to hook them all up to a monitor and keep switching hdmi inputs from the monitor(if you have just 1) in order to switch to other vms. With looking glass threw just 1 hdmi cable that connects to the gpu of your host OS you can run vms and be able to see them etc
@@enzoys No. You are talking of vGPU. That’s a complicated thing cause splitting the gpu probably needs Enterprise level GPU and other complicated stuff. About looking glass. Let’s say you have 2 GPUs in your system and with 1 GPU you run Linux as a Host Operating System. You will have that 1 GPU connected via hdmi to your monitor to be able to see it. Now let’s say with the 2nd GPU you run a windows guest vm. In order to be able to see the windows guest VM you must connect the 2nd GPU to a monitor again. So in order to be able to see your host and guest vm at the same time you either must have 2 monitors to have the vm in each of them or you will have to keep switching to the hdmi inputs in your monitor to either see the host or guest vm. That’s where looking glass comes in place to solve that problem. Instead of the 2nd GPU to show the display data from its hdmi output looking glass instead takes the display data and transfer’s it through the pci Express back to the motherboard and then through PCI Express again to the 1st GPU. So the 1st GPU that runs the host has the data both the host and guest so when you hook it up with an hdmi to 1 monitor you will be able to see The host and the guest vm as a window. The advantage of this over other methods are massive. Because this comes at latency that is not noticeable, the frames come uncompressed so no visual quality decrease. The only downside is that it takes ~10% GPU performance for that application to keep running. This is extremely useful especially in laptops cause they have 1 monitor. Without looking glass if you would make a vm with gpu passthrough you couldn’t be able to see it. Only if you would plug an external monitor.
@@hunhowshadow317 thanks a lot for this complete answer. And yea I just wanted to passthrough my gpu but still be able to use my linux, but the singles gpu guide I saw would log you off linux to use the vm and I think looking glass is the solution. Thanks!!
@@enzoys Yes single GPU passthrough completely “logs” you off Linux (it doesn’t actually you just can’t see it due to no gpu) and only be able to see after windows shut down
Just add another elif after line 35 like this: elif [ "$DISTRO" == "Arch" ] then pacman -Syu binutils sdl2 sdl2_ttf libx11 nettle fontconfig cmake spice-protocol make pkg-config gcc gnu-free-fonts or just change ManjaroLinux to Arch in line 30
That´s what I am looking for, too. Want to be able to use the same (main) GPU for Linux and Windows KVM gaming while using a second (weak) GPU for doing voice, browsing, etc on the host on second display.
@@PavolElsig the one guide I found for single gpy passthrough in theory needs to log you off linux to have the GPU on windows (I haven't tested it), I just wanted an option in which I could use my GPU on windows but still be able to use linux if I want to (and have the GPU back on linux when I shut down windows)
Strangely, I don't have any "Video Cirrus" in my virt manager, and I don't see this in add device menu. I see Display spice and Video QXL though, but I can't set "none" for "Video QXL" device if thats what it is...
Thank you for this guide, worked perfectly. If there is another ubuntu user the service wont work; is possible to create a service for a specific user that starts when login? or what do you thing is the best solution in this case?
While running the script I got this error I am using fedora 35 CMake Error at /usr/share/cmake/Modules/FindPkgConfig.cmake:603 (message): A required package was not found Call Stack (most recent call first): /usr/share/cmake/Modules/FindPkgConfig.cmake:825 (_pkg_check_modules_internal) displayservers/Wayland/CMakeLists.txt:55 (pkg_check_modules)
@Pavol Elsig Checking for module 'wayland-protocols>=1.15' -- No package 'wayland-protocols' found CMake Error at /usr/share/cmake-3.16/Modules/FindPkgConfig.cmake:463 (message): A required package was not found Call Stack (most recent call first): /usr/share/cmake-3.16/Modules/FindPkgConfig.cmake:643 (_pkg_check_modules_internal) displayservers/Wayland/CMakeLists.txt:30 (pkg_check_modules) my friend what should i do please can you simply explain why i got this error ? How can I fix.
@@PavolElsig i wanna switch completely to linux and a made vm for gaming a and windows things, and my school want us to use "LockDown Browser" and it detect im using a vm
Je-Sec It gets pretty complicated. I don’t know of any 100% way of doing it. You can check out this link to get you started: rayanfam.com/topics/defeating-malware-anti-vm-techniques-cpuid-based-instructions/
No luck, tried it on linux Ubuntu 22.10. it complains that no package 'libpipewire-0.3' was found even though it is installed then it stops. If you can help I would truly appreciate it!
I'm on Linux Mint 20.2 and I can't seem to compile and link looking-glass-client. This is what happns when I run "sudo ./lg.sh": 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. logname: no login name User: will be using Looking Glass on this PC. If that's correct, press (y) otherwise press (n) and you will be able to specify the user. y/n? And if I continue, I get: [100%] Linking C executable looking-glass-client [100%] Built target looking-glass-client chown: missing operand after ‘looking-glass-client’ Try 'chown --help' for more information. It seems that the problem has something to do with the owner, but I've never encountered this problem during installation before.
I have a new build and my questions are growing. Dual gpu build for gaming, video rendering, hashcat and also goprofusion. I understand the limitations on dual gpu, the software developer has to implement the dual gpu support. Amd 8core, 16 thread apu. Dual rx 580x. If I run a linux host on the apu, then pass the gpu's to the vm, will this setup be more likely to work out of the box with software that has dual gpu implemented?
mikey z It won’t improve compatibility with multi-GPU. Hashcat, rendering and such will work fine with any GPUs you throw at them. Games will work exactly the same as they would on bare-metal. The real benefit is that you can basically turn your singe PC into two or more workstations. You can have multiple sets of keyboards and mice and work on multiple things in multiple OSes at the same time.
How many graphics cards are needed? If there is only one monitor, should the monitor be connected to the discrete graphics card or to the core graphics card?
You do need two GPUs - one for the host and one for the guest. The monitor would be connected to the host GPU. Whichever GPU is not bound to vfio-pci, that’s your host GPU. Any of my GPU passthrough guides should explain how to do that.
hello there, pls help, I tried LG for many times, but never successfully to work. For my problem I think I install everything, but when run LG the LG-host showed nvbcf not work, it seemed there was no card in windows,so the LG-host cannot work, so my question is that how to config the vm(windows),really thanks
depends on how your motherboard groups the pcie devices, mine (x570 mpg gaming plus with the latest bios) have no problem doing that. search for iommu groups to find more
could you please do a tutorial on sound passthrough with Scream streaming it to pulseaudio by network and I think after that you'll have covered all my questions I ever had for pcie passthrough qemu émulation
@@ritzmat I gave up trying to use scream with looking glass due to a driver problem, although I can't remember it's been a long time. I used splice audio instead, pretty decent.
romi Looking Glass allows you to display guest OS in a window on the host’s monitor. But normally (without Looking Glass) you’d want two monitors or at least change inputs on the one monitor you have. It should work with RX 580 or AMD GPUs in general. Some AMD GPUs like Vegas and Navis have the reset bug, but the 580 should be fine.
@@PavolElsig ahhh i see so there is no need to have 2 monitors at all or switching displays on one monitor. How does the guest even operate doe if there is no GPU connected to it, given that the guest works in a window now instead on a monitor.
romi If doing passthrough, you’d assign a GPU to the guest. That way you get GPU acceleration and can play games. With Looking Glass, instead of seeing the output on a monitor connected to the guest GPU, the display output gets sent to the host and displayed on the host’s monitor in a window. Without connecting anything to the guest GPU, the resolution might be stuck on something low. For that reason it is recommended to connect a dummy plug or another input of your monitor. That fools the GPU into thinking that it is displaying to a monitor. In reality all output from the guest is being sent to Looking Glass.
thank you for the amazing video and for the scripts you're providing. just a note: I tried version B5.0.1 of looking glass and I get some errors when compiling it on Manjaro. I tried v B4 and it works. maybe this will help someone
not worth it. Its memory to memory copy, even if you can grab the buffer and send it over the network which maybe possible, the data is uncompressed meaning a 1080p stream will need something like 500mb/s for 60fps which again possible with 10G Lan. The one thing will get you is latency. I will sugest you to look over the moonlight project with a nvidia gpu.
@Pavol Elsig Hate to be the dumbest kid in class, but I am running your lg.sh downloaded from the Looking Glass helper githuib onto my Fedora 33 system and I fail at 100% with output: [100%] Built target looking-glass-client Please be patient Nothing to do libsemanage.map_file: Unable to open my-qemusystemx86.pp (No such file or directory). libsemanage.semanage_direct_install_file: Unable to read file my-qemusystemx86.pp (No such file or directory). semodule: Failed on my-qemusystemx86.pp! Any advice would be much appreciated, I'll try to manually compile the client in the mean time.
There WAS an issue with this, the QEMU RAM didn't have permission. I had to go in and delete existing policies of this type by doing: sudo semodule -r my-qemusystemx86 then sudo semodule -X 300 -r my-qemusystemx86 before running these commands manually directly: (as my regular local user with sudo) sudo ausearch -c 'qemu-system-x86' --raw | audit2allow -M my-qemusystemx86 sudo semodule -X 300 -i my-qemusystemx86.pp sudo setsebool -P domain_can_mmap_files 1
@@PavolElsig I have a feeling your script SHOULD work fine. Since, I really just stole the FEDORA if clause and reran it until it worked properly; best guess I somehow introduced a 400 policy prior to running your script in some way. After doing the policy fixing steps I outlined, everything has been working perfectly (I also had to fix my shared MEM buffer to being 128MB since I am doing 4k).
What is the point of this video? You alredy have 2 GPU's to start with... I don't have a 2nd GPU nor an Integrated one. That's why I needed Looking Glass. But you alredy have 2 GPU's and the VM's working... Is this just so you could use 1 monitor?
Yes, Looking Glass allows you to run GPU accelerated Windows VM in a window. Let’s say you wanted to play Windows games in Linux. You just spin up a VM open Looking Glass and play your game. You can alt-tab in and out of Windows back to Linux. You don’t need two monitors or two sets of keyboards and mice.
@@PavolElsig I know what it does, but you are setting it up with 2 GPUs to begin with. I can't pass the GPU thro before I set up Looking Glass but I alredy need to pass the GPU thro in order to set Looking Glass up... You enterd the system with a GPU alredy passed thro and then Installed Looking Glass, I can't do that. Should I set everything up then pass it thro or what?
@@faust9091 I set up Looking Glass with the default Spice and QXL. You don't even need two GPUs to use Spice and QXL. You just won't be able to take advantage of GPU acceleration if you keep using it.
Every new video of yours is like Christmas morning.
Thanks! Glad you find the videos helpful!
agreed
If I could I would give you more thumbs in and views. This is the titile truly speaks for itself
Thanks a lot!
I am glad you show how to install B2-rc4, but feel bad for you as you missed the final B2 release by literally a few days :(
gnif Thanks for all your hard work that went into making Looking Glass possible. I’ll put it into the description to let people know to use the B2 release
Thanks a ton! A note: LG B5.0.1 (and possibly earlier) needed the xpresent-dev package installed (via standard package manager) as well in order for the compilation to work.
Thanks for the video and for the script!
Everything worked, as it should be on PoP OS 20.04.
I always wanted to do hotswappable peripherals setup without a real physical KVM. And, here with this software, plus a monitor with two plugs, I am just happy ;D!
Thanks so much this was so easy to follow and it is incredible!!!
Great video ans script! Thanks for helping me with PopOS passthrough and LookingGlass. You are the best !
Hello!
Thanks for the brilliant guide. It has helped me to get Looking Glass up and running!
One quick note that could be helpful:
I had to buy a new GPU to replace my GTX660 that doesn't support UEFI. It seems as though the Nvidia 43 error cannot be avoided when using default SeaBIOS, so you must use the UEFI one instead (hence rendering old graphics cards unusable).
Cheers.
It should be possible to use Seabios with Nvidia GPUs. I have a GTX580 that I use every now and then. In my Windows XP video I did a couple extra steps to get a GTX 980 working under Seabios.
Just finished installing it on Ubuntu 20.04 and the last LG release. Work perfectly for my usage in 3440*1440 with 64M of memory. (passthrough GPU a GTX1080)
I'm using it to use my elgato 60S and game capture to play with my Switch. Since i use an external audio card, audio latency is less than 10ms avec according to the calibration of games like crypt of necrodancer, visual latency is under 50ms wich is fine.
Many thanks to you.
Glad to hear that everything works great for you. Awesome setup, btw
what are the advantages of using looking-glass? does this mean I can use the gpu inside my guest machine AND on my host machine?
Thank you so much! I tried so much and my guest GPUs video encoder is just too slow for most RDP tools but this is working great.
Awesome keep up the great work!
Thanks and I appreciate the feedback
What host gpu do you recommend for 1080p 144fps looking glass?
I get an unspecified extract error when trying to extract lookinglass in windows
Hi, why we need looking glass and not just boot up to guest OS?
because with GPU pass through you will have to hook them all up to a monitor and keep switching hdmi inputs from the monitor(if you have just 1) in order to switch to other vms. With looking glass threw just 1 hdmi cable that connects to the gpu of your host OS you can run vms and be able to see them etc
@@hunhowshadow317 what are the advantages of using looking-glass? does this mean I can use the gpu inside my guest machine AND on my host machine?
@@enzoys No. You are talking of vGPU. That’s a complicated thing cause splitting the gpu probably needs Enterprise level GPU and other complicated stuff. About looking glass. Let’s say you have 2 GPUs in your system and with 1 GPU you run Linux as a Host Operating System. You will have that 1 GPU connected via hdmi to your monitor to be able to see it. Now let’s say with the 2nd GPU you run a windows guest vm. In order to be able to see the windows guest VM you must connect the 2nd GPU to a monitor again. So in order to be able to see your host and guest vm at the same time you either must have 2 monitors to have the vm in each of them or you will have to keep switching to the hdmi inputs in your monitor to either see the host or guest vm. That’s where looking glass comes in place to solve that problem. Instead of the 2nd GPU to show the display data from its hdmi output looking glass instead takes the display data and transfer’s it through the pci Express back to the motherboard and then through PCI Express again to the 1st GPU. So the 1st GPU that runs the host has the data both the host and guest so when you hook it up with an hdmi to 1 monitor you will be able to see The host and the guest vm as a window. The advantage of this over other methods are massive. Because this comes at latency that is not noticeable, the frames come uncompressed so no visual quality decrease. The only downside is that it takes ~10% GPU performance for that application to keep running. This is extremely useful especially in laptops cause they have 1 monitor. Without looking glass if you would make a vm with gpu passthrough you couldn’t be able to see it. Only if you would plug an external monitor.
@@hunhowshadow317 thanks a lot for this complete answer. And yea I just wanted to passthrough my gpu but still be able to use my linux, but the singles gpu guide I saw would log you off linux to use the vm and I think looking glass is the solution. Thanks!!
@@enzoys Yes single GPU passthrough completely “logs” you off Linux (it doesn’t actually you just can’t see it due to no gpu) and only be able to see after windows shut down
is possible use looking glass for run two virtual machines with gpu passthrough at same time?
You should be able to do that, but you’d need to create a separate backing file for each
When I compile .sh file I get a cmake error message after No package 'libpipewire-0.3' found?
sadly it doesn't work on arch the script but still a great video for the other distro ! great job : D
Just add another elif after line 35 like this:
elif [ "$DISTRO" == "Arch" ]
then
pacman -Syu binutils sdl2 sdl2_ttf libx11 nettle fontconfig cmake spice-protocol make pkg-config gcc gnu-free-fonts
or just change ManjaroLinux to Arch in line 30
If a discrete graphics card is passed through to kvm, can the discrete graphics card be used on the host before kvm is turned on?
It is possible but it gets pretty complicated. If you look for single GPU passthrough guides, they deal with pretty much this exact issue.
That´s what I am looking for, too. Want to be able to use the same (main) GPU for Linux and Windows KVM gaming while using a second (weak) GPU for doing voice, browsing, etc on the host on second display.
@@PavolElsig the one guide I found for single gpy passthrough in theory needs to log you off linux to have the GPU on windows (I haven't tested it), I just wanted an option in which I could use my GPU on windows but still be able to use linux if I want to (and have the GPU back on linux when I shut down windows)
Strangely, I don't have any "Video Cirrus" in my virt manager, and I don't see this in add device menu. I see Display spice and Video QXL though, but I can't set "none" for "Video QXL" device if thats what it is...
Thank you for this guide, worked perfectly. If there is another ubuntu user the service wont work; is possible to create a service for a specific user that starts when login? or what do you thing is the best solution in this case?
I think you’d just have to adjust the permissions or ownerships of some of the files created when the script runs.
How well does this work for high refresh monitor.? I have a 144hz and I would like to game at 144hz.
You’d be better off connecting that GPU directly and just changing inputs. I’ve never tested it, but i think 144hz might be pushing it.
@@PavolElsig that’s works. Thanks
While running the script I got this error I am using fedora 35
CMake Error at /usr/share/cmake/Modules/FindPkgConfig.cmake:603 (message):
A required package was not found
Call Stack (most recent call first):
/usr/share/cmake/Modules/FindPkgConfig.cmake:825 (_pkg_check_modules_internal)
displayservers/Wayland/CMakeLists.txt:55 (pkg_check_modules)
@Pavol Elsig
Checking for module 'wayland-protocols>=1.15'
-- No package 'wayland-protocols' found
CMake Error at /usr/share/cmake-3.16/Modules/FindPkgConfig.cmake:463 (message):
A required package was not found
Call Stack (most recent call first):
/usr/share/cmake-3.16/Modules/FindPkgConfig.cmake:643 (_pkg_check_modules_internal)
displayservers/Wayland/CMakeLists.txt:30 (pkg_check_modules)
my friend what should i do please can you simply explain why i got this error ?
How can I fix.
Y there are two pci devices listed, are they pass through gpus?
I tried so many times but no image came out
can this be done with a ryzen cpu and a single gpu?
How to hide some windows programs from detection ?
Je-Sec What program are you trying to hide from detection?
@@PavolElsig i wanna switch completely to linux and a made vm for gaming a and windows things, and my school want us to use "LockDown Browser" and it detect im using a vm
Je-Sec It gets pretty complicated. I don’t know of any 100% way of doing it. You can check out this link to get you started: rayanfam.com/topics/defeating-malware-anti-vm-techniques-cpuid-based-instructions/
Your doing nice job helping to get virtual machine work, thx :)
Any chance to get looking glass working on single gpu passthrough?
No luck, tried it on linux Ubuntu 22.10. it complains that no package 'libpipewire-0.3' was found even though it is installed then it stops. If you can help I would truly appreciate it!
The looking glass docs say these are required:
apt-get install binutils-dev cmake fonts-dejavu-core libfontconfig-dev \
gcc g++ pkg-config libegl-dev libgl-dev libgles-dev libspice-protocol-dev \
nettle-dev libx11-dev libxcursor-dev libxi-dev libxinerama-dev \
libxpresent-dev libxss-dev libxkbcommon-dev libwayland-dev wayland-protocols \
libpipewire-0.3-dev libpulse-dev libsamplerate0-dev
I'm on Linux Mint 20.2 and I can't seem to compile and link looking-glass-client.
This is what happns when I run "sudo ./lg.sh":
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
logname: no login name
User: will be using Looking Glass on this PC.
If that's correct, press (y) otherwise press (n) and you will be able to specify the user.
y/n?
And if I continue, I get:
[100%] Linking C executable looking-glass-client
[100%] Built target looking-glass-client
chown: missing operand after ‘looking-glass-client’
Try 'chown --help' for more information.
It seems that the problem has something to do with the owner, but I've never encountered this problem during installation before.
I have a new build and my questions are growing. Dual gpu build for gaming, video rendering, hashcat and also goprofusion.
I understand the limitations on dual gpu, the software developer has to implement the dual gpu support.
Amd 8core, 16 thread apu.
Dual rx 580x.
If I run a linux host on the apu, then pass the gpu's to the vm, will this setup be more likely to work out of the box with software that has dual gpu implemented?
mikey z It won’t improve compatibility with multi-GPU. Hashcat, rendering and such will work fine with any GPUs you throw at them. Games will work exactly the same as they would on bare-metal. The real benefit is that you can basically turn your singe PC into two or more workstations. You can have multiple sets of keyboards and mice and work on multiple things in multiple OSes at the same time.
How many graphics cards are needed? If there is only one monitor, should the monitor be connected to the discrete graphics card or to the core graphics card?
You do need two GPUs - one for the host and one for the guest. The monitor would be connected to the host GPU. Whichever GPU is not bound to vfio-pci, that’s your host GPU. Any of my GPU passthrough guides should explain how to do that.
@@PavolElsig Thanks
hello there, pls help, I tried LG for many times, but never successfully to work. For my problem I think I install everything, but when run LG the LG-host showed nvbcf not work, it seemed there was no card in windows,so the LG-host cannot work, so my question is that how to config the vm(windows),really thanks
So, this is impossible for Win11 due to its mandatory to have secure boot enabled
Thank you!
wouldnt you have to passthrough the whole chipset if you plugged a gpu into it?
depends on how your motherboard groups the pcie devices, mine (x570 mpg gaming plus with the latest bios) have no problem doing that. search for iommu groups to find more
could you please do a tutorial on sound passthrough with Scream streaming it to pulseaudio by network and I think after that you'll have covered all my questions I ever had for pcie passthrough qemu émulation
Scream audio definitely deserves a new video. I'll look into it
@@PavolElsig thank you so much your channel is so helpful I really appreciate everything you've done for the KVM gaming community
@@ritzmat I gave up trying to use scream with looking glass due to a driver problem, although I can't remember it's been a long time. I used splice audio instead, pretty decent.
@@shanesreal Do you need unsigned drivers for it ? that's what I saw in an other video
I don't get it do you need two gpus ? i have two monitors connected to one gpu can i use this software ?
You need 2 gpus to use this.
Do you need 2 monitors running for this even after setting up Looking Glass? Or can you just have one monitor
romi You just need one.
@@PavolElsig thank you for replying so quick! I thought you needed 2. Does this work for the XFX Radeon RX580 too? Or just amd cards in general
romi Looking Glass allows you to display guest OS in a window on the host’s monitor. But normally (without Looking Glass) you’d want two monitors or at least change inputs on the one monitor you have. It should work with RX 580 or AMD GPUs in general. Some AMD GPUs like Vegas and Navis have the reset bug, but the 580 should be fine.
@@PavolElsig ahhh i see so there is no need to have 2 monitors at all or switching displays on one monitor. How does the guest even operate doe if there is no GPU connected to it, given that the guest works in a window now instead on a monitor.
romi If doing passthrough, you’d assign a GPU to the guest. That way you get GPU acceleration and can play games. With Looking Glass, instead of seeing the output on a monitor connected to the guest GPU, the display output gets sent to the host and displayed on the host’s monitor in a window. Without connecting anything to the guest GPU, the resolution might be stuck on something low. For that reason it is recommended to connect a dummy plug or another input of your monitor. That fools the GPU into thinking that it is displaying to a monitor. In reality all output from the guest is being sent to Looking Glass.
thank you for the amazing video and for the scripts you're providing.
just a note: I tried version B5.0.1 of looking glass and I get some errors when compiling it on Manjaro. I tried v B4 and it works. maybe this will help someone
Same. Looks like cmake errors out with some issue with X11 on versions >= B5.
I confirm , it does not work with B5...
How can i contact with you ? I have some question for antidetect OWASP
You can ask me right here. I have never used antidetect OWASP, so I don’t know whether I’ll be able to help you.
@@PavolElsig do you have telegram ? I want to talk more with you
Hi man, it is possible to use looking glass for connect remote kvm machine ?
not worth it. Its memory to memory copy, even if you can grab the buffer and send it over the network which maybe possible, the data is uncompressed meaning a 1080p stream will need something like 500mb/s for 60fps which again possible with 10G Lan. The one thing will get you is latency. I will sugest you to look over the moonlight project with a nvidia gpu.
nope, unless you have a shared RAM device
Does this need 2 gpus?
04:36 You know you can add that in the virt-manager GUI too?
Hello!!!
Please update the script for opensuse tumbleweed
thank you very much
great tutorial!
Thanks
You should really update the script. Fedora doesn't install all the dependencies anymore to run the script
does this still work?
@Pavol Elsig Hate to be the dumbest kid in class, but I am running your lg.sh downloaded from the Looking Glass helper githuib onto my Fedora 33 system and I fail at 100% with output:
[100%] Built target looking-glass-client
Please be patient
Nothing to do
libsemanage.map_file: Unable to open my-qemusystemx86.pp
(No such file or directory).
libsemanage.semanage_direct_install_file: Unable to read file my-qemusystemx86.pp
(No such file or directory).
semodule: Failed on my-qemusystemx86.pp!
Any advice would be much appreciated, I'll try to manually compile the client in the mean time.
Wait, the client DOES exist in the location it should, so maybe this is an entirely ignoreable error. Gonna continue on.
There WAS an issue with this, the QEMU RAM didn't have permission. I had to go in and delete existing policies of this type by doing:
sudo semodule -r my-qemusystemx86
then
sudo semodule -X 300 -r my-qemusystemx86
before running these commands manually directly:
(as my regular local user with sudo)
sudo ausearch -c 'qemu-system-x86' --raw | audit2allow -M my-qemusystemx86
sudo semodule -X 300 -i my-qemusystemx86.pp
sudo setsebool -P domain_can_mmap_files 1
@@lutzleipold5206 I’ve never tested it on Fedora 33. I have no idea whether it’ll work on it.
@@PavolElsig I have a feeling your script SHOULD work fine. Since, I really just stole the FEDORA if clause and reran it until it worked properly; best guess I somehow introduced a 400 policy prior to running your script in some way. After doing the policy fixing steps I outlined, everything has been working perfectly (I also had to fix my shared MEM buffer to being 128MB since I am doing 4k).
@@lutzleipold5206 Thanks for pointing it out. I’ll take a look at this.
its out dated and does not work anymore
I didn't know host GPU performance matters......oh no
THE CANE
too complicated
lol u know csgo is native to linux
What is the point of this video? You alredy have 2 GPU's to start with... I don't have a 2nd GPU nor an Integrated one. That's why I needed Looking Glass. But you alredy have 2 GPU's and the VM's working... Is this just so you could use 1 monitor?
Yes, Looking Glass allows you to run GPU accelerated Windows VM in a window. Let’s say you wanted to play Windows games in Linux. You just spin up a VM open Looking Glass and play your game. You can alt-tab in and out of Windows back to Linux. You don’t need two monitors or two sets of keyboards and mice.
@@PavolElsig I know what it does, but you are setting it up with 2 GPUs to begin with. I can't pass the GPU thro before I set up Looking Glass but I alredy need to pass the GPU thro in order to set Looking Glass up... You enterd the system with a GPU alredy passed thro and then Installed Looking Glass, I can't do that. Should I set everything up then pass it thro or what?
@@faust9091 I set up Looking Glass with the default Spice and QXL. You don't even need two GPUs to use Spice and QXL. You just won't be able to take advantage of GPU acceleration if you keep using it.
If you need help setting up GPU passthrough, I have guides for that too. I even have a guide for single Intel GPU passthrough with gvt-g.
@@PavolElsig So you can set it up with out the 2nd GPU. Thanks, I thought It was a requierment. I think i know what to do now