Puppy Linux Discussion Forum

norgo wrote: Sun Jan 15, 2023 1:54 pm

...
Important hint
qemu_gui is included in rootfs-skeleton by default

PPM offers a Qemu_Control_Center-0.7 package.
This package contains a not working qemu_gui script.
Do not install this pet because the working qemu_gui script would be overwritten!
qemu_gui_07.jpg

Agreed.

But PUP development team or PUP scripting members have not resolved it. The change-author of that script is not on the forum to resolve its field issues, after its last change.

I took a look awhile ago and do not have the skills to understand coding fo a screened program to resolve the problem...although I think I see the areas to resolve. Maybe there's an older working version around somewhere.

The GUI is helpful as it makes generating the VM visually simple for the simple things it does. The OLD PUP Guides on Google reference that feature of the QEMU package released in WoofCE repos and Puppyland PETs. So in the past, it did work.

Thanks Norgo

qemu-ready

is a little helper application to check whether your computer is ready for usage of qemu or not.
What is this application doing?

detecting the CPU manufacturer
loading the needed kvm kernel modules
checking the capability of cpu virtualization
checking whether virtualization is enabled or disabled in BIOS

after these tests you get a message if your computer is ready to run qemu

information
for detection of virtualization status (enabled/disabled) the script is using the "rdmsr" utility.
This very small utility is able to read the MSR (cpu register).
The binary is included in the package.

How to use?
Install the package.
open a terminal and type thats all.

An EXCELLENT utility that is NEEDED for PC that test/run forum distros via a QEMU virtual PC. This utility far exceeds the limited scope of the good earlier works the community helped me with several months ago.

This tiny PET, it is hoped that most forum developers include as one of the many OOTB utilities that we find useful considering that use of the Linux kernel carries this feature for a decade or more.

It clearly alerts the user that his PC is capable of the performance feature the kernel give the PC when a VM is used by the user.

Above post shows what happens for PCs capable of the kernel performance ability while below shows an Intel PC that does NOT have the ability to support the kernel feature. Thanks for this "QEMU_Ready" utility for checking and enabling the kernel where possible.

P,S, There is NO system penalty when the Linux kernel feature is present in the running system.

Hello @dimkr. Is your 'working' GUI, mentioned by @norgo above, now contained in either the QEMU packaging for Vanilla repos or a PET somewhere for PUPs (WoofCE)?

Thanks in advance

For FossaPUP64 and F96 and derivatives: @Jasper has released the latest QEMU v8+ as a PET.

My Virtual machines are running at 100% of the responsiveness of the Host Fossa.

Enjoy

Hi @rockedge; I remember you mentioning building kernels with KVM built-in.

Have you been able to run QEMU KVM without needing to modprove the KVM support before running QEMU on your systems that have the kernel?

Curious

@Clarity Yes some of the kernels I compiled earlier all have kvm support built in.

In F96-CE_4 (or any earlier version) a useful tool is the package cpu-checker which includes the command kvm-ok for checking the system status for virtualization

The current KLV-Airedale Void Linux kernel 6.1.20_1 seems also to have KVM support built in.

Clarity wrote: Mon Apr 24, 2023 3:10 am

For FossaPUP64 and F96 and derivatives: @Jasper has released the latest QEMU v8+ as a PET.

My Virtual machines are running at 100% of the responsiveness of the Host Fossa.

Enjoy

It looks like the PET package referenced has been removed from that post/topic but....
rockedge has compiled QEMU-8.0.0-rc3 (7.2.93) in F96-CE_4 as a PET, download link at the following forum post.
viewtopic.php?p=86333#p86333

@rockedge @dogcat
Thanks very much for the refresher. Just wanted to check to know which distros have the kernel built-in vs which did not. Your kernel update is not present in older forum distros.

Also, think the command you suggest looks fine.

But the forum PET by @norgo here; "qemu-ready" seems small, rather inclusive and detailed. It goes a step further that if the kernel module is not loaded, it will

• activate it for the user

• or alert the user of BIOS/UEFI need to turn on by the user.

AS 'kvm-ready command' has been tested extensively on my old test PCs. Wanted to insure it works with testing new kernels containing the KVM support that members use.

Thanks again.

rockedge has compiled QEMU-8.0.0-rc3 (7.2.93) in F96-CE_4 as a PET.......

Yes and I am using it right now on a F96-CE_4 as host machine running KLV-Airedale-rc15 and KLV-Airedale-rc16

qemu-ready ( build script )

Because qemu-ready is using rdmsr ( msr-tools ), it's not possible to create a pet that will work for all CPU architectures, distros and versions.
If you are interested in qemu-ready, you can build the qemu-ready pet from source.
The only thing you have to do is to load the DEVX sfs.
After that run the attached script.
This script will download the msr-tools sources from smokey01.com server and compile the msr-tools.
The rdmsr binary is needed as already mentioned and the only binary which is going into the pet.

Usage at own risk

See here for an example of using qemu to boot a frugal install (of a FR/KL firstribit creation) directly from hard disk without making an image or iso of it:

viewtopic.php?p=126064#p126064

frugal2qemu.sh utility based on my immediately above post that will auto-boot an already made frugal install as a qemu virtual machine (read only):

viewtopic.php?p=126355#p126355

Made to work with any of FirstRib/KL, Puppy Linux, DebianDog, FatDog, and EasyOS, but can easily be reconfigured to work with other frugal install or full install type distros.

For many users unaware, the 'qemu_ready' utility by @norgo is 'still' required to automatically & easily turn on KVM before QEMU use on the host recently rebooted.

Otherwise, you must manually turn on with no error checking status given on each reboot.

This is because most developers are not including the support in the OOTB kernels built within their distros.

@rockedge is an exception as he has provided kernels with the support enabled.