QEMU 8.0 Released With 32-bit x86 Host Support Deprecated

[phoronix]

Phoronix: QEMU 8.0 Released With 32-bit x86 Host Support Deprecated

QEMU 8.0 is out today as the newest feature release for this processor emulator that plays an important role in the open-source Linux virtualization stack...

https://www.phoronix.com/news/QEMU-8.0-Released

[Nozo]

Since the old 32-bit has been deprecated, how long it will take for the new 32-bit (x86-64-v1) hosts to be deprecated from QEMU?

[ayumu]

Remember when qemu was all about emulating (or virtualizing) everything on anything?

Commercial interests sure have perverted the project.

[stiiixy]

Remember when qemu was all about emulating (or virtualizing) everything on anything?

Commercial interests sure have perverted the project.

That's what I was thinking. I've had to use it a few times to get stuff like MacOS 6 era accounting data files slowly upgraded through the OS and package versions to the modern format. Which was inferior but basically useless because tax tables and whatnot.

[kiffmet]

ayumu Noone is preventing 32bit users from just continuing to use the 7.X series or creating a fork that reverts the commits removing x86_32 host support. AMD64 was introduced 19 years ago - there has been plenty of time to switch over.

I also don't think that the overlap of people running a 32bit x86 CPU and needing (not wanting, needing!) to emulate i.e. the latest RISC-V + all ISA extensions on a Venn-diagram is particularily big (or for that matter, existent) either.

[CommunityMember]

Remember when qemu was all about emulating (or virtualizing) everything on anything?

Commercial interests sure have perverted the project.

Support for 32-bit hosts has been deprecated because of lack of resources. You are free (encouraged even) to contribute your organizations resources to maintain 32-bit host support should you desire it.

[Developer12]

Remember when qemu was all about emulating (or virtualizing) everything on anything?

Commercial interests sure have perverted the project.

The way I see it, deprecating the ability to *host* on a 32 bit machine is less of an issue than, say, deprecating the ability to emulate old 32 bit systems. The standard response is "host on newer hardware" but you don't loose any overall ability to emulate older systems to do whatever you need to do with them.

I'd probably be ok with rigorously maintaining the ability to emulate every system from the past and present, but only do so on relatively modern systems and maybe some notably powerful+popular older ones.

Maybe I just don't adequately appreciate some aspect of being able to emulate everything on literally *everything* else though.

[Quackdoc]

Remember when qemu was all about emulating (or virtualizing) everything on anything?

Commercial interests sure have perverted the project.

I myself am conflicted, on one hand I agree, but on the other it makes a lot of sense, sadly qemu just doesn't have enough people working on it to become a "mega project" so to speak.

[brad0]

Remember when qemu was all about emulating (or virtualizing) everything on anything?

That has never been a thing.

[kpedersen]

I wish they would deprecate the Gtk GUI instead. This is just taking up needless developer resources. It is severely incomplete and useless. Anything interesting needs to be done via the qemu monitor or via command line flags. Not to mention on most distributions it drags in a bunch of random dependencies, even for the Gtk hooks. It is an example of a GUI just for the sake of saying it has a GUI.

I doubt there are commercial vendors using it and a home user probably only think that they use it due to lack of knowledge of the alternative SDL, VNC, stdio/serial.