Running with mouseoff produced this:

DOSBox SVN, CPU speed:
SOUND.CFG
UYVETIKT
36 File(s)
12 Dir(s)

D:\NASCAR>nr1999r
DOS/4GW Professional Protected Mode Run-time Version 2.01a
Copyright (c) Tenberry Software, Inc. 1996
Copyright (c) 1994-1996 Papyrus Design Group, Inc. All rights reserved.
Version 2.00
Pentium-class machine.
Running from D: drive
UniVBE driver not present
Unable to read options.cfg file. Using defaults.
Unable to read race.cfg file. Using defaults.
Unable to read controls.cfg file. Using defaults.
Digital Sound
Port 220 IRQ 7 DMA 1
Conf irmed with SB16 8 ST
587008 Conventional
Triple Buffering
Warning: Did not find a match in mode table
NASCAR Racing 2 requires a mouse. Please load your mouse driver.

It's definitely video related.

sounds like it corrupt something in memory and invokes this error, i guess

Well to track the mouse, the current resolution has to be known. The new rendition code probably doesn't update something.

Another test:
[EDIT] Skip this
https://1drv.ms/u/s!As-dKk-N73dSmCifLGR2NbOvzSCZ?e=AqsBjy
MD5: c463fbd3e46d442d32e59a32a6959544

Run this instead:
https://1drv.ms/u/s!As-dKk-N73dSmClVhXB3PuDy3ZF7?e=t8EAZ6
MD5: 899a97e9b0a685109d99e0a8ca942399
[/EDIT]

This should dump a few more calls. [EDIT] And test the FIFO transfer mechanism. This isn't documented very well[/EDIT].

There's something weird about VESA SetMode call 0x4F02. The mode passed isn't in the list passed by call 0x4f00 and queried by 0x4f01.

The mode numbers passed are: 0x807F and 0xC12E.

corrupted
executed in win98 got txt file which in attachment.

execution under dos , doesnot saves txt file, mode don't returns, everything on artifacts.

https://fex.net/s/zpe5zom

the video at fexnet is execution under dos, if you have vlc player you can press 'e' to look every frame, its still output something.

the command setmode.exe | more don't show anything at all

EDIT: the previous one you said to skip, works well.

I think sending a V_FIFO_CONTEXT_INIT to the Verite was a mistake. This time I'm sending a V_FIFO_RESERVED1, which should be a noop equivalent.

I thought of sending a harmless V_FIFO_SQUARE to draw a square, but I don't know how to swap buffers yet (or even create a primary surface (i.e. I need it's address in verite space).

https://1drv.ms/u/s!As-dKk-N73dSmCqUOEFnZR0z3yHg?e=VOgpTb
MD5: f348d8691bb2d3a0ac9a42efb0d30c49

[EDIT] There should be another text file SM.txt, which I need.[/EDIT]

The results of modes 0x807F and 0xC12E. with VESA call 0x4F06 are weird (with BX=1 to query pixel resolutions).

For one of them I get a width of 640, linebytes 1280 and a max scanlines height of 0xCCC (3276)
and for the other width 1280, linebytes 1280 and max scanlines 0xCCC.

again win98 in attachment
dos in video https://fex.net/s/dcyclat

Apologies for the delay, Turbo C++ is giving me problems. In particular the mode list query no longer works if the Int 10h call is from a subroutine:

[EDIT]
Skip this:
https://1drv.ms/u/s!As-dKk-N73dSmCvkVPh2OTJjMke5?e=wRfd5f
MD5: 84735e4c89fa5d3e302b886e7306904b
and skip this:
https://1drv.ms/u/s!As-dKk-N73dSmCwn0VtK-UZwW-6r?e=npm1q2
MD5: c9bcd7350cbbe7e16e713366a2a35470

Third time's the charm:
https://1drv.ms/u/s!As-dKk-N73dSmC0GcQ-LoIQoyE_8?e=sgdw4A
MD5: 5149e3bfc2c69bb1ce336cc3fcb85a9e

Everything should be working now.
[/EDIT]

Redirected output is what I would like OR you may end up with screen corruption or a lockup. You may just see a red rectangle in there (I seriously doubt this is going to work).

third, draws artifacts for a some seconds, and them corrcetly returns to dos, with upped hz of monitor.
https://fex.net/s/evkpbno

And now this:

https://1drv.ms/u/s!As-dKk-N73dSmC4ZHxEHeXu96S45?e=fX3GHz
MD5: 3e911d5d945b8bf72176227906d381d8

https://fex.net/s/5mombpb

seems nothing new. monitor switched to 70hz on exit before it was 60 hz, everything else intact, e/g/ setmode text same as before

There was no corruption right?

[edit] yes there was.

no corruption , just capture card can;t capture 70hz, everything looks correct

What's your desktop resolution and what's the res while fifo's running?

my monitor recognizes it as:
textmode 720x400@70 (this somehow can be captured)
one sec of fifo is 640x480@60
than return to text mode 720x400@70 (which can't be grabbed by capture card)

Three more fifos:
https://1drv.ms/u/s!As-dKk-N73dSmC-8TeBnx4VzEw58?e=OHVVnX
MD5: eb3d2534c9dd8f42f7ccd61beded830e

Things to check:
1) Stability/No lockups
2) Corruption
3) Red Square appearing on screen.
4) 60 Hz on exit.

And two more

https://1drv.ms/u/s!As-dKk-N73dSmDDGoWAerflLB6Fn?e=YGxOhr
MD5: 6354278d455b7ceb812dac0c8b002861

That's all folks (at least for tonight).

I lied here're two more:
https://1drv.ms/u/s!As-dKk-N73dSmDHFuQy-1YD84M_J?e=IsZapl
MD5: 492639ab321ac65913882fdb1495b94a

With FIFO5 and 6 there should be a delay. If this works, FIFO6 *might* render a red square.