86Box: Emulator of x86-Based Machines | Hacker News Reader

86Box: Emulator of x86-Based Machines

github.com

97 pointstosh38 comments

I am a big fan of 86box for my retro PC hobby. Some neat things I've used it for:

* Log virtual POST codes to a text file for a physical motherboard I'm troubleshooting to see where it's hanging up

* Install an old OS to a .img file which I can then extract and pull bit-perfect files from

* Create a virtual hard drive with matching CHS values to a real hard drive, install an OS, then write the image to a real disk and boot it on real hardware

* And of course, play around with old hardware, software, and games in fairly accurate simulated way :)

>Log virtual POST codes to a text file for a physical motherboard I'm troubleshooting to see where it's hanging up

I went as far as simulating defects by modifying individual hardware blocks in 86Box source code, recompiling, and checking resulting POST code because that seemed faster than carefully tracing BIOS :)

"Modified 86Box to break RTC always returning 0 or 0xff. POST_7 freeze." https://www.vogons.org/viewtopic.php?p=1041079#p1041079

That's awesome, rasz! It's amazing how 86box allows us to peer into hardware we otherwise couldn't easily access. 8)

I was blown away how easy and fast whole process was. I posted about my plan at 14:33 and results at 16:55. I even commented at the time how "setting up build env and compiling 86Box was a breeze".

> Log virtual POST codes to a text file for a physical motherboard I'm troubleshooting to see where it's hanging up

That sounds awesome but my brain can't follow today.

What did the setup look like here?

My brain's looking at a test motherboard, then at 86Box (which would presumably run on your main system), and I can't figure out how to connect the dots.

Definitely! I've been meaning to write a post about it.

I had a 486 motherboard that would hang somewhere during POST. I have a POST card that displays the current code from the ISA bus, but even with my phone camera in slow-mo mode, I wasn't able to capture the full list of codes - the value on the bus changed faster than what the POST card could output.

I turned to 86Box and found a very similar 486 board with the same chipset. There's an option to install a virtual POST card and show the codes in the application's status bar, but this also went by too fast for me to see what was going on.

So I downloaded the source, found the file for the POST card, and edited it to write the codes to the console as well. Then I fired up the virtual board again and produced a complete log of the POST codes during a cold start.

Together with the complete list from 86Box plus the codes from the real board, I found it was completing many steps but hanging on "chipset initialization". I still have some work to fix it, but it at least gave me some insight I wouldn't have otherwise had. :)

That is crazy and awesome that the virtual POST card (brandnewsentence) modeled the real environment so closely. Feels obvious that it would line up - once proven, haha.

I would have been yelling at an Arduino (and possibly a shift register) wired to the 7-segment LEDs... (I don't do hardware :D but I would've been convinced it was necessary)

How does this compare with blink [1] ?

[1] https://github.com/jart/blink

86box seems to be the successor of PCem, a GUI Emulator that is capable of emulating full Hardware (incl. Settings and BIOS/firmware) via GUI selection. I use it for old games, because emulating of 3dfx and similar is also possible in an easy way.

Maybe this can be done with blink, but probably not by selecting a few dropdowns and clicking some buttons...

Pcem is still being actively developed, even though Sarah Walker said she wasn't working on it anymore.

Successor or fork?

I am not sure about the details, but I seem to remember it was not a very amicable fork.

It's a fork, going back some years (2017 or so).

It's instructive to look at the activity level of pcem (their forum, and their github account: https://github.com/sarah-walker-pcem/pcem/commits/dev/) versus 86box ( https://github.com/86Box/86Box/commits/master/ )

86box has a vibrant community around it (pop in the discord some time!) which PCEM really doesn't as far as I'm able to tell.

Yes, there is a lot of bad blood -but I would suggest taking any stories that you hear (but did not witness yourself) with a huge grain of salt.

There is also a video about PCEm with a reference to 86box from LTT:

https://www.youtube.com/watch?v=mWEE2RJj3YI

Found this pretty interesting how to use in practise.

Not at all. Blink according to themselves is:

> blink is a virtual machine that runs x86-64-linux programs on different operating systems and hardware architectures.

Where 86Box is, according to themselves:

> 86Box is a low level x86 emulator that runs older operating systems and software designed for IBM PC systems and compatibles from 1981 through fairly recent system designs based on the PCI bus.

Blink seems to be closer to qemu-user or box86/box64[0].

0. https://box86.org/

wow, the screenshots attached to each release on github are deeply nostalgic

https://github.com/86Box/86Box/releases

This is an excellent piece of software - use it almost every day. It’s the best way to run basically any operating system from the pre-USB era. It includes a very good emulation of the IBM PC _jr_, a whole host of other ISA, VESA Local Bus, MCA, and PCI based machines. It also has CPU modes good enough to run troublesome programs that don’t function in DOSBox or dosemu2.

I do occasionally run into stability issues, especially with 3D games. The emulation is not 100% perfect.

This is a nice project. I use it to emulate BeOS on an Apple Silicon Macbook.

Is this needed because recent x86-64 processors / hardware are not able to run older 32bits binaries anymore?

Getting any legacy OS to run on modern hardware is very hard. For example, in my case, BeOS has no EFI boot capability, has a serious RAM restriction, doesn't really like anything past a P4, and has no drivers for anything but legacy hardware (like, no modern storage drivers - IDE only, no SATA unless there is legacy BIOS emulation.) A lot of this is able to be worked around with various hacks, but with 86Box, it will run the OS as is it was on the actual hardware, so it just makes it a no brainer. You can then also use it to emulate old PC hardware on other processor architectures - like ARM.

It runs on other CPU architectures and it emulates older PCs which include other hardware besides the CPU.

Ah yes, they have ARM binaries indeed in the download section. I got trapped by the "minimum requirements": Intel Core 2 or AMD Athlon 64 processor or newer, which made me think it only ran on x86

Did you mean 16-bit binaries?

Strictly speaking, no. It's primarily developed (as I understand it) for cases (gaming, mostly, but some hobby/dev use) where emulating hardware accurately (think timings) is desireable.

In a sense it's also a history preservation project, though I doubt they would characterize it as such themselves. They don't do the documentation that pcjs does; but I think it's similar in spirit, if not in implementation specifics.

I should clarify that when I say that, I don't mean documentation about the emulator (which is coming along nicely), but documentation about the computers being emulated. I was specifically thinking of things like this ( https://www.pcjs.org/machines/pcx86/compaq/portable/ ) when I was writing that comment.

AFAIK that's a software limitation, not a hardware one (yet).

Technically it is an hardware limitation, amd64 in long mode has only vestigal support for memory segmentation (enough to implement Win32/ELF i386 TLS ABI and that's it), so you cannot reasonably run either 16b protected mode or vm86 tasks in 64b OS on that platform.

To add to what userbinator said, you can have 16-bit code segments in long mode. I've played around with this a bit, and made a minimal proof-of-concept "DOS emulator": an INT 21h from the 16-bit segment causes a SIGSEGV, which is intercepted by 32-bit code and turned into the corresponding Linux INT 80h syscall. All running on a standard 64-bit kernel with no dependency on any external libraries.

(I'm not planning to publish this, and it's very minimal: the only supported syscalls are read, write and exit, and the program needs to have exactly two segments, CS and DS=initial SS as defined in the EXE header. But it should be theoretically possible to make something more elaborate that can run at least a few "well-behaved" normal DOS programs, as long as they don't do any arithmetic on segment registers.)

What's no longer available in long mode is the virtual 8086 mode, for that you'd need to use either the more modern virtualization features (with "unrestricted guest" support), or a software emulator.

so you cannot reasonably run either 16b protected mode or vm86 tasks in 64b OS on that platform.

I don't know who has been spreading FUD (likely MS, just like they did with "32-bit Windows can't use more than 4GB of RAM") but that's not true.

https://www.dkia.at/en/node/180

There's another article I can't find at the moment which discusses this in a lot of technical detail, making references to using the "unrestricted guest" virtualisation feature on CPUs that have it (quite new at the time) which essentially runs 16-bit code as a VM, or classic modeswitching that jumps between 16, 32, and 64-bit mode as needed. Nonetheless, it is quite irritating that despite the hardware being perfectly capable of 16-bit, 32-bit, and 64-bit operation, the designers made it harder than necessary to do so.

In the absence of VM extensions, which the earliest AMD64 CPUs did not have, once the CPU is in long mode, the base address of segments other than FS and GS doesn't work, it must be 0.

That's not such a problem for 32-bit segments whose base is normally 0 anyway (flat 32), but most 16-bit protected mode Windows code expects the linear address of its code and data segments to be something other than 0, and even might expect to be able to use multiple data segments with different base addresses.

So, sure, you can run 16-bit protected mode code with a base address of 0, but that's not very useful.

Also there's no virtual 8086 mode when in long mode to support real mode code.

With VM extensions it's possible to virtualize the CPU in regular (not long) mode, at which point the virtualized CPU will indeed support non-zero segment base addresses and virtual 8086 mode.

In theory this can be done with an extremely light virtual machine layer, but the entry/exit to/from the 16-bit code, or even for that matter any code needing to use 16-bit data segments is going to require more than it would if the CPU weren't in long mode. And this is assuming you have virtual mode extensions. If you don't then you need a CPU emulator.

My guess is that Microsoft felt it just wasn't worth writing all of this extra code needed to do this. There's a big difference between technically possible and makes even remote economic sense to do it.

When the manual says that the segment bases are ignored in long mode, it is only referring to 64-bit code. As far as I remember, the terminology goes like this:

    64-bit OS, running 64-bit user code -> "long mode"
    64-bit OS, 16/32-bit user code -> "compatibility mode"
    16/32-bit OS -> "legacy mode"

Also I've personally tested that it works, without any virtualization required (see sibling comment). Wine also does it in order to run 16-bit Windows programs, or at least it used to at one time - IIRC, there was some discussion on the Linux kernel mailing list about it.

Intel dropped the "x86s" project to focus on a more collaborative approach with amd and others, so it's probably been shifted a few years out anyway.

It allows you to emulate MS-DOS or other legacy 16/32-bit operating systems on any supported hardware, such as ARM.

See also, the main site: https://86box.net/

On PC emulation/DOS related, you used to have a FreeDOS 1.2 tgz for DOSEMU and I can't find it nowhere.

Not to be confused with https://box86.org/

How does it handle serial to TCP connection? I'm using DOSBox now, you can configure a serial port to turn into a TCP connection.

Reply on news.ycombinator.com