Is it true that Nvidia BIOS does not support VGA BIOS state save/restore functions?

I would like to get confirmation that Nvidia VGA BIOS does not have functional VGA BIOS state save/restore functions.

The problem is that these functions seem to be just dummies in the Nvidia BIOSes, doing nothing.
This results in problems with suspend/resume, because this depends on these BIOS functions do their work.

Nvidia devs, can you please confirm this, or otherwise explain why attempting to restore state does not to what it does on any other brand graphics card?

And, is there an alternative way to make saving/restoring state work on Nvidia cards via BIOS?

Are you trying to suspend and resume at the console?

I’m not sure about the details of the legacy VGA save / restore functions, but suspend & resume is only supported when the nvidia-modeset driver is initialized. Typically that means either running an X server or nvidia-persistenced, or by loading nvidia-drm with the “modeset=1” parameter.

If you have something that’s making low-level VGA BIOS calls during suspend & resume, please disable it. Things like that can get in the way of the driver’s resume handling.

Are you trying to suspend and resume at the console?
On computers without xorg, there is only console.
Suspend/resume then always results in no signal from the Nvidia card.
This indicates that the state save/restore BIOS functions do not do what they should do.

If there is no way to get around this problem, the information on the FreeBSD hardware compatibility page should be updated accordingly. (-> https://wiki.freebsd.org/Graphics#NVidia_.2F_GeForce_Graphics )

In addition to that, the lack of these BIOS functions break suspend/resume for Nvidia cards on computers with xorg.

The developer of the vesa.ko driver explained this issue:

The root cause is NVIDIA does not really support VESA BIOS
re-POST and save/restore state calls. It just checks few
bits and returns immediately if they are set/clear. I
have a feeling that they intentionally did it to make
reverse engineering harder. On top of that, they don’t
provide any information to improve the situation.[/code]

Here is the link: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=224069#c9

Personally, I think Nvidia should pay for, respective sponsor implementing the workarounds necessary to fix the issues resulting in selling cards with incomplete VGA BIOS, if the above statement is true.

Okay. No answer is also an answer :(

Feeling forced to migrate all workstations to AMD cards.
With these, suspend/resume works fine and reliably in all modes - out-of-the box, without any additional configuring.

Had to find out that the statement I made in my last post is wrong.
The evaluation AMD/ATI card I was testing appeared to work OK.
But not for long.

There were quite a lot of issues with the Radeon driver:
Driver uploading wrong microcode to card.
Resume failing after a few suspend/resume cycles.
And more small things one not immediately notices.

So I installed the Quadro card back and retired the AMD thing on the test server, which is run in text mode only.

Conclusion:
If you need good graphics, Nvidia currently is the only choice.
If you do not use graphics mode, a cheap AMD/ATI card might be okay.