NVidia Video Driver Shutdown Fix

[schwups]

Machine (NV 7800GT PCIE), driver 82.16: After I corrected "45 4E" to "15 4F" the protection error is gone  , but the power supply doesn't shutdown and the cursor remains  . And there was no Scandisk after shutdown. A subsequent manual scandisk didn't find any errors  .

Edited July 19, 2015 by schwups

[loblo]

 

Is it possible, that in this case is a difference between ME and 98SE? Loblo, does the patch work for you?

 

 

I don't know, I applied the patch to 82.69 which I have installed and rebooted once since without issues but I I haven't try to shut down and haven't a series 6 or 7 card in the machine at the moment.  

 

Tomorrow I'll probably try to replace the FX5200 I have installed with the "new" 7950GT 256MB AGP I bought a little while ago but haven't tested so far and hopefully  I will be able to tell more, that is if this card work at all.

Edited July 18, 2015 by loblo

[Drugwash]

Machine (NV 7800GT PCIE), driver 82.16: After I corrected "45 4E" to "15 4F" the protection error is gone  , but the power supply doesn't shutdown and the cursor remains  . And there was no Scandisk after shutdown. A subsequent manual scandisk doesn't found any errors  .

 

Yes, that was to be expected. The code doesn't appear to work as it should (at least on your machine), but additionally there was that calculation bug that jumped to a wrong location leading to the protection error. The 'red fix' was only for that issue, for the rest is up to Mr. Loew.

 

However, I've been working on an automated patcher based on the code pattern presented in first post and so far it appears to work correctly, but we'll have to wait for some additional fixes.

 

While I'm here, does anybody know for sure which driver version is the very first to exhibit the shutdown bug so I could set the lower limit for the patcher? Also, which is the lowest card version that exhibits the bug in conjunction with the drivers? I'd like to set some kind of frame so as not to unnecessarily patch driver versions other than the ones that do need patching.

 

Of course, all that if Mr. Loew agrees to the patcher - after all, it's not my code.

Edited July 18, 2015 by Drugwash

[schwups]

I'm not sure, but I think driver 81.85 should be the first affected driver, because 77.72 doesn't exhibit the shutdown bug. 77.72 supports 6xxx series. I seem to remember the 7xxx series require 82.16 or 82.69.

 

I never heard that one of 5xxx series is affected, which corresponds to my experience, too. Geforce 6xxx series is affected, but I never used a 6200. I'm not sure about this card.

[Drugwash]

OK, so I guess I should settle for a minimum driver version of 80.00 and card version 6000 (round figures).

Well, probably not tomorrow since it's my birthday, but sometime next week I'll add those checks. Thanks!

[rloew]

@ schwups: There may be a calculation error in the relative jump for 82.16 (if my logic is correct ). Please try to replace the values in red below:

31FCAC: 3C 32 75 02 B0 52 89 44 24 04 E9 45 4E E5 FF

to

31FCAC: 3C 32 75 02 B0 52 89 44 24 04 E9 15 4F E5 FF

For what it's worth I've used the XVI32 editor, simple and handy (click left pane, type hex values, click 'Save').

You are correct. The destination of the Jump is 174BD0 not 174B00. The Font setting for my DOSBox makes 'D' and '0' look similar.

@Schwups: Your problem may be different from mine. Resetting from the hung screen left me with Dirty Flags in my Hard Drive.

[rloew]

@Schwups:

Are you using ME or 98SE? I only tested it on 98SE.

You mentioned a Cursor. I assume you mean a DOS Cursor. I get a Black Screen with no Cursor. I don't think the resolution is being changed back to 640x480 either.

If you want to experiment, try changing just the 32 to one of the following, boot Windows, Shutdown (not Reboot or Boot to DOS), then try the next.

24 thru 2F, 34, 35 or 36.

[LoneCrusader]

I tested both the 81.98 and 82.69 patches.

The 81.98 patch works fine on a SOYO 875 Dragon 2 Platinum with a 6200GS.

I did notice that the "Windows is shutting down" screen never appears, it just goes black from the desktop and then shuts the machine off after a second or so.

The 82.69 patch works fine on an Intel D875PBZ machine with a 7950GT.

Same notation about the "shutting down" screen.

I have also seen the "blinking DOS cursor" screen before, along with a screen that reports "unable to write to device AUX" or something like that. Both of these situations kept the machine from shutting down, but neither of them left the system "dirty" or caused SCANDISK to run on the next boot.

[dencorso]

Since you've provided patches for almost all the preferred drivers I have a request: would you please provide it also for the 77.72?

Here are the official links to it <ENU> or <International>, both are signed, but the signature in them has expired already, which is not a problem because the source is reliable... 

[rloew]

I already Patched 77.72 as I used it in some experiments.

I did not post the Patch because as far as I know 77.72 is not susceptable to this problem.

[dencorso]

I already Patched 77.72 as I used it in some experiments.

I did not post the Patch because as far as I know 77.72 is not susceptable to this problem.

 

Well, it all depends on how old one's nVidia card is: please do re-read this, this and (maybe) this. So the patch to 77.72 may, in fact, be useful.

[rloew]

I already Patched 77.72 as I used it in some experiments.

I did not post the Patch because as far as I know 77.72 is not susceptable to this problem.

Well, it all depends on how old one's nVidia card is: please do re-read this, this and (maybe) this. So the patch to 77.72 may, in fact, be useful.

None of your links actually say that the 77.72 Driver has a Shutdown Problem with any Card.

Just in case someone finds a use for it. The corresponding Patch for the 77.72 Version is as follows:

144: 14 -> 23

2D394E: 2E 1A E7 FF -> C2 05 00 00

2D3F14: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 -> 3C 32 75 02 B0 52 89 44 24 04 E9 5D 14 E7 FF

Drugwash probably could use this to verify his Patcher at the low Version end.

The code in this portion of the File seems to have little if any changes between 77.72 and 82.69.

[dencorso]

Thanks a lot!

I remember having tested it. But I cannot find my notes for that, and I'm not sure I had already started using the RAM Limitation Patch, when I did those tests... and now I know I was very near the limits of what can be done with the A7V600-X until starting to use the the RAM Limitation Patch, so that I cannot guarantee I'd have the same results today, in case I repeated those tests. But I've added them to my to-do list, anyway.   

[schwups]

@Schwups:

Are you using ME or 98SE? I only tested it on 98SE.

You mentioned a Cursor. I assume you mean a DOS Cursor. I get a Black Screen with no Cursor. I don't think the resolution is being changed back to 640x480 either.

If you want to experiment, try changing just the 32 to one of the following, boot Windows, Shutdown (not Reboot or Boot to DOS), then try the next.

24 thru 2F, 34, 35 or 36.

- I use ME. I install 98SE for rare test purposes only.

 

- Yes it's the DOS cursor on black screen.

[rloew]

I did not test it with ME so I have no idea if it would work for me either. There may be more System Messages beyond 36 to test.