FreeBSD on legacy nvidia hybrid laptop

A

AmerXz

Hi I'm a newbie to FreeBSD and I have an old laptop with intel 4600hd and nvidia gtx 765M (should work with nvidia 470 driver)

I tried to do some research but most methods uses nvidia-drm-kmod (which doesn't seem to be available for legacy drivers) or using nvidia-hybrid-graphics which is deprecated

also I don't have any option in the laptop's bios to disable hybrid mode

my question is

is there any way to configure my hybrid setup or at least disable intel gpu from xorg config ?

my device is msi GE60 2OE from around 2013
 
loveydovey said:
So no wayland with that card?
Click to expand...
Maybe it would depend on what compositor is used.
Starting with 470 series of drivers, graphics/egl-wayland should be functional.

So if you're lucky enough and the compositor used supports EGLStream-based Wayland external platform, it may possibly work without graphics/nvidia-drm-*-kmod on Wayland. But such a compositor would be limited. Not sure what compositor actually support EGLStream.
 
SirDice said:
Logs appears to indicate that it's working on the Intel driver. Are you not getting a picture on the screen?
Click to expand...
Xorg is indeed working but when doing
xrandr --listproviders
It only show intel but not nvidia

Also nvidia-smi works

When trying to use vkcube it tries to run on nvidia gpu then vkcube crashes with segmentation fault
 
AmerXz said:
I asked claude ai about the situation and it recommended me to connect a monitor to the hdmi connector to make nvidia gpu work

Conversation here
Click to expand...
It mutually depends on how the notebook is "physically" designed whether the suggestion works or not.

As far as I've found before (cannot give you a link, I've lost track with it as it was 5+ years ago), at least some notebooks had "external" HDMI and/or DP port directly attached to NVIDIA dGPU, but not to internal display panel.

ThinkPad P52 (at the moment) alone had configuration to disable iGPU via UEFI firmware and let NVIDIA dGPU to drive internal display panel directly.
(To be honest, it seems that internal display panel TP P52 has multiple DP ports and attaching one for iGPU and another one for dGPU with auto selection of active port.)

Some manufacturers / models clearly stated that NVIDIA dGPU can be used via Optimus only.

And all others shows no clear information about the above.
And maybe some og them work indirect rendering without PRIME offloading (let dGPU as if some monitor is connected regardless actually do so or not, and tranfer the results of dGPU rendered image into VRAM area of iGPU, managed and transferred by softwares like x11/virtualgl).
 
You must log in or register to reply here.
Back
Top