Hi,
Here's a new version of MSKVBEF7 that seems to work at least with Quake 1. It modifies the NumberOfImagePages field in the ModeInfoBlock structure.
More info in this thread:
Old+Modern videocards pure DOS benchmarking- which one is fastest?, need your numbers + analysis, 320x200 to 1600x1200!

Thanks for your effort, it works, even for Blood and DND3D (and sure it will for all Build-based games too 😀
Good work. Now it is nice playable at 1600x1200 mode at 40-55 FPS.
You can add this utility to 1st post for other random visitors, maybe change topic name to mark we have a workaround now.

BTW I also tested vid_wait 1 option in quake and it turn on vsync in non-page flipping modes - I got FPS capped at 60Hz
For Ruthan: my time demo benchmark - 67,2 FPS @1280x1024 with MTRRLFBE, machine below exc. VGA is GTX670.

Thanks for your tests. I have modified topic's title and added download link to first post as you have suggested.
@RayerR: Now something completely different. We had a broken conversation earlier about MTRRLFBE not working under protected/V86 mode with some systems. Can we continue this, please? Original conversation:
X58/i865/V880 - Yamaha7x4/AurealV1/2 pure Dos7.1- compatibility list/research/ultim. drivers configs, WIP- gurus needed

@Edit: LFB WC performance problem under v86 mode confirmed by RayeR:
https://www.bttr-software.de/forum/forum_entry.php?id=15770

When trying an A64+ 939 machine for high-res 98/DOS gaming the build engine games (Duke3D/Blood) gave me issues on my FX5900 Ultra and Geforce 4 Ti4200 -- I had to drop down to a Geforce 3 for them to work well without -- though the built-in FPS counter still 'lied' at certain resolutions. Is the bug mentioned here part of this or does this only apply to modern GPUs?

No, in your case it is something different. This thread is about new GPUs GTX 5xx+, older was not affected.

I just tested with patch with My MSI Geforce 980, if it would fix that high resolution text modes bug with File Wizzard Filemanager and its still broken, so there has to some other bugs with these new Nvidia cards. Its not nothing special its just default FW mode, broken from Geforce 730 GDR5 above.

Picture is completely broken.

And what mode FW use? Does it have some config that you can select different mode that works OK? You can try ask FW author what modes he use and what VESA library/what functions to focus wehere the problem is...

I dunno details, but you can download my drivers and utils package, here is FW folder with File Wizzard with same settings which im using.
https://www.dropbox.com/s/4g8snw60ouf1jzl/DRI … 181113.zip?dl=0

I already, tried to find details about FW author to try to find out, why when i try to start first program after its start and boot, it crashes and for second time run its all fine.. Its some hungarian guy with pretty generatic name, i wasnt able to find some working contact.

ruthan wrote:

I just tested with patch with My MSI Geforce 980, if it would fix that high resolution text modes bug with File Wizzard Filemanager and its still broken, so there has to some other bugs with these new Nvidia cards. Its not nothing special its just default FW mode, broken from Geforce 730 GDR5 above.
Picture is completely broken.

Hi,
I do not think this problem has anything to do with VESA. AFAIK NVIDIA never supported VESA Text modes at all or at least finished supporting VESA Text modes long before Kepler/Maxwell (my Geforce 6600 GT does not support any VESA Text modes at all either). The missing support is also reflected by FWIZARD since all the extra high resolution modes are disabled (it seems strange at first but green means mode is disabled and yellow means mode is enabled in FWIZARD).
This problem is triggered by a non-standard VGA Text mode when 8px character width is requested with 720x400/90x25 resolution. So this problem can be considered a bug but definitely not a VESA BIOS bug. And since this problem is related to a non-standard VGA resolution I consider this a not so serious issue.

I didnt check details about patch, i hoped its simply fix main Dos problems for these new Nvidia cards.

I have now Geforce 6600 GT in same machine, i thing that its from Gigabyte and its fine, im using FW as default FM, so i tested it with 50+ cards, ATI / AMD / Matroxes /S3/ 3Dfx.. all in my Dos videocards performance lists.. and Geforces from 4MX up to 730 DDR3 PCI-E all these new are ok, only these Nvidia with PCI-E gen 3 arent working with FW.
I never tested it with Geforce 4xx - 6xx line because i dont have these cards. I also dont have Radeons HD 7000 and newer.

i hoped its simply fix main Dos problems for these new Nvidia cards.

No, the patch is very specific in this sense as it can be read in its description:
'MSKVBEF7 is a TSR that masks/hides VESA VBE Function 07h - Set/Get Display Start. It can somewhat cure the rather serious VESA BIOS bug of Nvidia Kepler/Maxwell/Pascal series that can cause flickering/skipped frames under DOS programs that use hardware page flipping for double/triple buffering.'

I have now Geforce 6600 GT in same machine, i thing that its from Gigabyte and its fine, im using FW as default FM, so i tested it with 50+ cards, ATI / AMD / Matroxes /S3/ 3Dfx.. all in my Dos videocards performance lists.. and Geforces from 4MX up to 730 DDR3 PCI-E all these new are ok, only these Nvidia with PCI-E gen 3 arent working with FW.

Hi,
I have not said this non-standard VGA mode is also buggy with Geforce 6600 GT. I have just said that it's not buggy on Geforce Kepler/Maxwell because of VESA Bios bug since VESA Text modes are not supported by NVIDIA VESA Bios at all.

Here are some pictures what I'm talking about:
1.a. If VESA text modes are supported then you get something like this in FWIZARD:

1.b. And something like this in AIDA 'Graphics -> SVGA VESA Text Mode' section:

2.a. If VESA text modes are not supported then you get something like this in FWIZARD (as I said green means not supported):

2.b. And in AIDA:

In case of both Geforce 6600 GT and Geforce Kepler/Maxwell you get something like the 2. case: VESA Text modes are not supported at all.
But yes, Geforce 6600 GT and Geforce Kepler/Maxwell are different in the way that 720x400/90x25 non-stanard VGA mode is supported by Geforce 6600 but not by Geforce Kepler/Maxwell.
But this mode is not a standard VGA Text mode so the support is not 'demanded' by VGA standard:
https://en.wikipedia.org/wiki/VGA-compatible_ … mmon_text_modes
The bug is triggered on Geforce Kepler/Maxwell when you select any modes represented by the 4th yellow 'X' column in FWIZARD.
I mean these ones:

PS:

Its some hungarian guy with pretty generatic name, i wasnt able to find some working contact.

I'm also Hungarian so I can say he does not have a 'general' name as far as Hungarian names are concerned. On one hand 3 names are not so general in Hungary (the typical is 2) and overall 'Gulyás Antal Dénes' is rather distinguishable. I have found some reference and it seems he was an active member of the community of late famous 'Enterprise 128' era in Hungary, but I also could not find an active contact.
But it seems you can reach 'Laci' at http://www.ep128.hu/ and maybe he can help since one of his pages refer to a program written by 'Gulyás Antal Dénes' and Laci's site seems to be still active (last modification date is 2019.03.30.)
http://www.ep128.hu/Ep_Games/Leiras/Tetris.htm

Thanks for detail analysis of problem. So its rare case.
On paper, by the book, its not standard implementation, but if all other videocards manufacturers support these modes (and they do), its de facto standard mode, even older Nvidia cards supported them.

To author thanks for research, i will try to ask middleman, i only can tell that when i googled him, i find more than 2,3 guys with same name, im not sure its long time, but i maybe because i wasnt successfull with 3 names, only used first and last name not middle name.

Just don't wonder that newer cards has worse VESA implementation as it is common thrend to strip everything legacy down and focus on new stuffs - UEFI. During years Nvidia dropped various VESA features like refresh rate setting for CRT monitors, 8x14 font, etc. so we have to live with it or use some older VGA. I personally don't use this hires text VESA modes so I didn't meet this problem. As page flipping patch works for me I can live with GTX670 quite well. What I'm missing is accelerated driver for Win98 but it's diff. story...

RayeR wrote on 2018-12-22, 17:42:

Beside in Ruthan's thread Old+Modern videocards pure DOS benchmarking- which one is fastest?, need your numbers + analysis, 320x200 to 1600x1200! I found he stated:

Geforce 730 difference, i have two G730 cards, one with GDRR3 and one with GDRR5 - i though than difference would be only in memory, but card with GDRR3 has not blinking problem and working with VIA 880 chipset, GDRR5 card has blinking problem (again Q1 only Q2 same resolution is not affected) and not working with VIA no picture at all + broken Text modes in FileWizzard. It could be manufacturer or bios difference, slower if Gigabyte, faster is Zotac, its i also has 4 GB of VRAM what is of course useless on such card..

So this prove that not ALL modern nvidia cards are afected by this bug and there may be a chance to find also some 6XX BIOS that is not affected and could be flashed to GTX 670 after some mods. Or if I get a dump of good video BIOS from GTX 730 I can disassemble it and try to look for code of VBE func. 07h and compare it with bad version. Good function then may be used as a TSR or direct BIOS insertion patch...

Has anyone made any further progress on this issue?

Also, there are 96 core and 384 core versions of GT730 and versions with DDR3 and versions with GDDR5 to further complicate things. A GT730 card might not be a good card to test with due to it's split personality issues.

As absurd as it sounds, there are Fermi (pre Kepler) models of the GT730 as well, just like the GT630 which was often a relabeled GT430 Fermi card. They will have 96 shaders in a GF108 core, with DDR3. The GDDR5 models are always Kepler, but DDR3 models have a chance of being Fermi.

Most likely that's what you've got there. Load up GPU-Z and it should tell you.

Sadly, the Fermi models are horrible in comparison to Kepler. Less than half the pixel fill rate, less than half the FLOPs, almost twice the power consumption... Just bad.

I got some reports of DOS users about newer nvidia cards GTX 1070 and RTX 2060:
http://www.bttr-software.de/forum/board_entry … id=15762#p16525
and unsurprissingly newer nVidias sucks in DOS even harded - they messed up VGA register compatability. So avoid anything newer when you wand to use native DOS. I hope GTX 9x0 are not worse than GTX 670, also they are last with WinXP support so I'm not interested in newers anyway.

FWIW I tried to run the application under XP x64 with a Maxwell Titan X, but it seems the test application only supports x86.

I think I'll get another hard drive soon, and then I can try to run it under x86.

What application do you exactly mean? Utils discused here are targetted for real DOS, not Windows.

Is there some simple Vesa / Vga test, with some simple working/ failed summary? It would be good start, better than test it on game basic.

Other question is we are sure that all this problems are only videocard related, that other variables arent in the play? For example how about Notebook with optimus - intel + nvidia? I have notebook with Geforce 2060 too.. and unless of was generation with Intel + Nvidia 10xx there is not simple switch in bios to select which card is primary in modern OS both cards are always detected and in nvidia panel is some switch for which application which gpu should be used.

I'm still learning, but I was using VBEVSTST.EXE

Falcosoft wrote on 2017-12-13, 02:56:

I have written a little test (VBEVSTST.EXE attached, source code included) that can be used to discover if a card is affected. After staring the program in DOS/Win9x/XP you just have to press the arrow keys.

This seemed to imply it should at least run in XP and identify if the card was affected, but I can understand x64 may have required a separate build or change to source.