vmwgfx_dri.so is part of the libgl1-mesa-dri package. I think the only question is why your install does not have it. I've checked mine and a number of colleagues machines and they all have it either in /lib/dri or /usr/lib/dri. The only common point is they are all non-VMs. Is your Ubuntu a VM?I think you should be able to fix this by installing the libgl1-mesa-dri package (sudo apt install libgl1-mesa-dri)
I installed the libgl1-mesa-dri, but the problem is not solved.
Now there is vmwgfx_dri.so,
I copied vmwgf.so to $PVLIB_HOME/GGA/Mesa/Linux64_GCC-9.3/lib path, and set the environment variable export LIBGL_DRIVERS_PATH=$PVLIB_HOME/GGA/Mesa/Linux64_GCC-9.3/lib
At this time, vmwgfx_dri.so has the problem of undefined symbol: amdgpu_query_video_caps_info
Set the environment variable export LIBGL_DRIVERS_PATH=/usr/lib/x86_64-linux-gnu/lib. At this time, both vmwgfx_dri.so and swrast_dri.so have the problem of undefined symbol: amdgpu_query_video_caps_info.
It seems that the version of vmwgfx_dri.so provided by libgl1-mesa-dri is too new compared to the Fast Models Third Party IP package
What is the version of mesa provided by the Fast Models Third Party IP package, and how can I get the corresponding version of vmwgfx_dri.so?
Thanks!
I seem to have solved this problem, but I am still not successful in booting android.
I downloaded and compiled AOSP, branch android-12 and lunch aosp_arm64-eng, and made myimage.img with ramdisk.img and system.img in the android out folder according to Fast Models User Guide (arm.com)
Here is my start command:
./FVP_Base_RevC-2xAEMvA --plugin=../../plugins/Linux64_GCC-9.3/Sidechannel.so
-C DEBUG.Sidechannel.interceptor=$PVLIB_HOME/GGA/reconciler/linux-x86_64/gcc-9.3.0/rel/libReconciler.so
-C pctl.startup=0.0.0.0
-C bp.secure_memory=0
-C cache_state_modelled=0
-C cluster0.NUM_CORES=1
-C cluster1.NUM_CORES=0
-C bp.smsc_91c111.enabled=1
-C bp.smsc_91c111.mac_address=auto
-C bp.pl011_uart0.untimed_fifos=1
-C bp.secureflashloader.fname=fvp_bl1.bin
-C bp.flashloader0.fname=fvp_fip.bin
-C bp.virtioblockdevice.image_path=image/myimage.img
But it stuck in the following interface：
I think there may be a problem with the android image produced, and the fast models user guide is too brief. How can I correctly build an android12 image and Device Tree Binary that can be used by fast models? What files do I need to prepare to run android under fast models?