Try to install DOS on modern laptop

>Try to install DOS on modern laptop
>"Unable to control A20 Line! XMS Driver not installed"

I thought x64 was supposed to be backwards comparable

Attached: 1547320370974.png (223x226, 5K)

Other urls found in this thread:

en.wikipedia.org/wiki/A20_line#A20_gate
os2museum.com/wp/the-a20-gate-fallout/
twitter.com/SFWRedditImages

The issue is with the frimware, not with the architecture itself. You can run virtual machines with DOS just fine on any modern x86 machine.

Attached: 65658.gif (530x360, 158K)

Install the driver then, idiot.

If you want to control A20, you need to be in Protected mode, so you need to install a Dos Protected Mode Interface driver.

If you can't manage that; just use dosbox like the rest of the morons.

>x64

Attached: 1532500988166.jpg (480x360, 16K)

>t. retard
at least you used the correct picture in your thread

you forgot to load himem.sys retard

That error happens when I try to load himem or xmsmmgr

It may be blocked by BIOS, check it for "Gate A20" option or
device=himem.sys /A20CONTROL:OFF

Wow an actually helpful reply on Jow Forums, thanks buddy

>backwards comparable

>lose the hma
That's the price for having an Intel, though.
en.wikipedia.org/wiki/A20_line#A20_gate

a20 gate is always enabled on newer intels so hma is accessible

Then it's the worst of both worlds.
I'll wager (but would like to be proven wrong) DOS won't load into the HMA if HIMEM.SYS (or equivalent) can't control the A20 gate.
It will break software that expects wraparound (a fair bit of early software): os2museum.com/wp/the-a20-gate-fallout/

DOS is loaded into HMA by HIMEM.SYS (DOS=HIGH), so there is no problem

>himem.sys loads
>does its usual probing around in xms
>finds it can't control gate a20
>moves the os there anyway
Not buying this.

Attached: 1545751796696.jpg (680x680, 36K)

> himem.sys /a20control:off
> already know it not needed to enable a20 gate and start loads dos to hma directly

>if i can't control gate a20 its open
That's a heck of an assumption for a memory manager to make. My Google-fu is weak though, and not turning up much on the subject. Empirical testing is unpalatable because this very machine is my only post-Haswell Intel.

It's not an assumption it's your direct command. Without this option it will throw the OP message and will not try to use HMA

Just install it in a Virtual Machine dumbass.

You didn't have an Intel 364 processor back in the day?

actually someone on this board who knows stuff and even helps. What the fuck is happening? Is this real?

The message is "unable to control A20", not "A20 is already on", though.
But having said that, HIMEM.SYS does appear to be able to detect whether the A20 gate is on or off (makes sense, it would be trivial to write to memory at (say) 0000:ff00, and see if it's mirrored at ffff:ff10):
/A20Control:OFF HIMEM only takes control of the A20 line if it is off when Himem.sys loads.

It happens frequently, user. Far more often in /sqt/ and suchlike, though.

Thanks for that, user. But what about programs depending on address wraparound? They'll still break without gate A20 being controllable.

You're too obsessed with the A20 gate, dude. It was at the end of the day a 286 workaround. On 386 or better, you can get into V86 mode (with EMM386 or equivalent), you can map memory how you see fit - including emulating the A20 gate entirely in software, I suspect.