Javascript x86 emulator

[mrstobbe]

Very nice... someone (Fabian) definitely had fun.

[mikegonta]

Have you seen this?
http://copy.sh/v24/
great fun, and a quick way to try out simple OS's

Is it supposed to be able to handle 32-bit mode or is it limited to real mode? ... It puts up that message within a few seconds with
MikeOS32 and my own OS.

Appears to be fixed and works fine. MikeOS32 is here.

[Muazzam]

I think, it does not like my OS. Still, it's a good approach.
Does it support SSE instructions?

[BrightLight]

It runs my OS properly, though the font doesn't seem to work.
It's a little odd that the processor doesn't have a model, though. Not exactly efficient, but definitely very cool, seeing as how we can now test our OSes from smartphones and such.

[SpyderTL]

My OS seems to run pretty well on this emulator, but only when booting from the hard drive. The CD and floppy boot loaders both lock up. I guess that's what I get for writing my own boot loader...

Still, pretty impressive.

[SpyderTL]

If anyone is interested, I found this list of online JavaScript emulators.

http://www.cambus.net/emulators-written-in-javascript/

The emulator in the original post is on this list.

[embryo2]

If anyone is interested, I found this list of online JavaScript emulators.

http://www.cambus.net/emulators-written-in-javascript/

A lot of work and a lot of time were spent. Is the JavaScript emulation so useful (or was it)? Or is it something that students make for some processor related courses?

[max]

A lot of work and a lot of time were spent. Is the JavaScript emulation so useful (or was it)? Or is it something that students make for some processor related courses?

No offense, but why are you bringing *anything* into question, whatever its about? They are clearly just doing it for the lulz and to show that its possible, probably also to show off. Same reason why people on here write their own OS, are they very useful?

[SpyderTL]

I think they are useful, because it gives me yet another test bed to test out my OS without having to buy any more hardware.

For instance, the one linked in the original post supports VirtIO Filesystem, which I've never seen before, and am currently researching.

Speaking of which, does anyone have any documentation for this PCI device? I've found some information on the 9p file system, but it only covers the command packets, not the "hardware" involved...

[Kevin]

The old pre-OASIS virtio specs seems to contain a bit on 9p devices: https://github.com/rustyrussell/virtio-spec

[SpyderTL]

I actually already have the VirtIO 0.9.5 spec PDF file. I didn't realize that it included the 9p transport subsystem id. But reading through that document again, I don't see any other information other than the ID. Maybe I'm just missing it...

I may try to look through the source for that VM in the original post, if I can find it.

Usually, the VirtIO communication protocol requires a VirtIO specific header in each packet, but maybe not in this case. Maybe you just write 9p packets directly to the VirtIO buffers. I may just try that when I get a few minutes.

[Kevin]

Yes, the old PDF specs that I can find are too old. The source in the repo I posted does have an appendix for virtio-9p, though.

I didn't read it, but it was rather short, so I guess apart from some initialisation you really just put the 9p packets (if that is what they are called) into the virtqueue.

[embryo2]

No offense, but why are you bringing *anything* into question, whatever its about?

No offense, but I don't understand what do you mean when you use the word "anything". Is it about why I ask the question? Or is it why I think there could be any other intent except the lulz? I really can't figure out what to answer here.

[SpyderTL]

Yes, the old PDF specs that I can find are too old. The source in the repo I posted does have an appendix for virtio-9p, though.

I didn't read it, but it was rather short, so I guess apart from some initialisation you really just put the 9p packets (if that is what they are called) into the virtqueue.

I didn't see this in any of the documents that I found. Can you post an exact link, or instructions on how to find that appendix? Thanks.

[Kevin]

It's pretty much the only file with real contents in the repository that I linked above, but here it is: https://github.com/rustyrussell/virtio- ... o-spec.lyx

It's probably easiest to just clone the whole repository and build a PDF document, but if you really want to read it in the source file, look for Appendix J, starting at line 9906. It doesn't say much more than what should be in the header files and "When the driver wantes to send a 9P request it places the descriptor of the PDU in the queue. It will be processed by the device according to the 9P protocol specifications (available at http://ericvh.github.io/9p-rfc/) and returned to the driver."