Pi Zero format boards that are supported.

you could try to boot it and see if panfrost works
but it does not build so
there are mismatches in various structs between 14.2 and drmsubtree
 
So
covacat said:
there are mismatches in various structs between 14.2 and drmsubtree
Click to expand...

so,maybe it will work with 14.1 if not even lower. Infact I've suddenly thought about to try 13.5. That's because the date of the last change made to the repo is 3 years ago.
 

kernel142drm.zip

drive.google.com drive.google.com

14.2 kernel build
it took a billion fucks but it the end it built. im not sure it will work in any way because some structure members / apis changed and i fixed them without much research / fully understanding everything.
just rename your current kernel
mv /boot/kernel /boot/ok-kernel
unzip this,zip in /boot (will create the /boot/kernel dir)
 
Thanks. I imagine that I should also install the "Panfrost userspace driver" ?

Code: 
# git clone https://gitlab.freedesktop.org/mesa/mesa
# mkdir build && cd build
# meson .. . -Degl=enabled -Dgles1=enabled -Dgles2=enabled -Ddri-drivers= -Dvulkan-drivers=-Dgallium-drivers=panfrost,kmsro
ninja
 
there is no 14-CURRENT anymore. newer pinebook pro image (the official ones) most likely do not have drm built in. mostly an efi fb
 
covacat said:
drmsubtree was never in an official arm build
os if you download an image you wont have accelerated graphics
Click to expand...

But why ? they created it for nothing ? Anyway I will try to upgrade 14.0 to 14.2 using the freebsd-update script,maybe it will persist even on the newer versions.
 
ok. But I want to achieve the maximum performances available for the "phone",maybe it will be more like a phablet, that I'm building.
 
Not sure about arm/arm64, but for i386 and amd64, in-tree drm was implemented natively, thus, become toooooooooo far more behind with Linux ones supported by hardware manufacturers as of quite limited resources, thus, drivers for X/Wayland counterparts does not work on in-first-hand-market GPUs with the lack of kernel part supports.

So, IIUC, FreeBSD project decided to dispose native ones and implement LinuxKPI layer to use Linux implementations, making dual-licensed-with-BSD and independent components could be in base and other parts to be in ports (graphics/drm-*-kmod).
 
I've "burnt" the FreeBSD-14.0-CURRENT-arm64-aarch64-PINEBOOKPRO.img file to the sd card and I tried to boot it using the rockpro rk3399 and the radxa zero 3W on my big monitor via HDMI. Unfortunately it didn't turn on in both the cases.

What I plan to do is to write the Jared Mcneill UEFI bootloaders to the sd card. On the UFS / root partition I will copy the content of the FreeBSD-14.0-CURRENT-arm64-aarch64-PINEBOOKPRO.img file.
 
covacat said:

kernel142drm.zip

drive.google.com drive.google.com

14.2 kernel build
it took a billion fucks but it the end it built. im not sure it will work in any way because some structure members / apis changed and i fixed them without much research / fully understanding everything.
just rename your current kernel
mv /boot/kernel /boot/ok-kernel
unzip this,zip in /boot (will create the /boot/kernel dir)
Click to expand...

_WhatsApp Image 2025-05-31 at 21.43.34.jpeg
 
Can't replicate this,because every time I swap the kernel dir,that error appears and it does not disappear even reverting the change (I mean,simply using cp and mv). I should dd again an old image and it takes 1 day (on Linux).
 
Code: 
Disk /dev/sdg: 238.42 GiB, 256003538944 bytes, 500006912 sectors
Disk model: STORAGE DEVICE
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: gpt
Disk identifier: E8E394A4-37BD-11F0-A6A4-002590EC5BF2

Dispositivo     Start      Fine   Settori   Size Tipo
/dev/sdg1       32768    135167    102400    50M EFI System
/dev/sdg2      135168 483844095 483708928 230.7G FreeBSD UFS
/dev/sdg3   483844096 499998719  16154624   7.7G FreeBSD swap

# dd if=/dev/sdg | pv | dd of=RockPro-RK3399.img bs=65536
dd: error reading '/dev/sdg': Errore di input/output

13792+0 records in
13792+0 records out
7061504 bytes (7.1 MB, 6.7 MiB) copied, 2.92741 s, 2.4 MB/s
6.73MiB 0:00:02 [2.30MiB/s] [  <=>                                                                 ]
0+2231 records in
0+2231 records out
7061504 bytes (7.1 MB, 6.7 MiB) copied, 2.92802 s, 2.4 MB/s

PS : without declaring bs=65536 I don't see the input / output error.
 
You must log in or register to reply here.
Back
Top