Hi, I’m the Toradex AI Assistant.
I might make mistakes or occasionally get you wrong. Our support team will step in if needed.
Based on your detailed description of the TIOVX/vision_apps firmware issues with BSP 7.3.0 on Aquila AM69, I can provide guidance on the correct approach to resolve the firmware and carveout address problems.
Solution for TIOVX/Vision Apps Firmware
To properly build and install the TIOVX/vision_apps firmware for the Aquila AM69, you need to follow the machine learning setup process that includes the TI Edge AI layers (1)(2).
Required Setup Steps
- Download the TI Layers: You need to obtain the TI Edge AI layers using their provided script (2):
$ cd /tmp
$ git clone https://git.ti.com/git/arago-project/oe-layersetup.git yocto-build
$ cd yocto-build
$ ./oe-layertool-setup.sh -f configs/processor-sdk-analytics/processor-sdk-analytics-11.00.00-config.txt
(2)
- Copy Required Layers: Copy the necessary TI layers to your Toradex BSP directory (2):
$ cd sources
$ sudo cp -a meta-arago/ <toradex_bsp_yocto_dir>/layers/
$ sudo cp -a meta-clang/ <toradex_bsp_yocto_dir>/layers/
$ sudo cp -a meta-edgeai/ <toradex_bsp_yocto_dir>/layers/
$ sudo cp -a meta-tisdk/ <toradex_bsp_yocto_dir>/layers/
(2)
- Update bblayers.conf: Include the Toradex compatibility layer and TI layers (2):
BBLAYERS += " \
${TOPDIR}/../layers/meta-edgeai-custom \
${TOPDIR}/../layers/meta-arago/meta-arago-distro \
${TOPDIR}/../layers/meta-arago/meta-arago-extras \
${TOPDIR}/../layers/meta-arago/meta-arago-test \
${TOPDIR}/../layers/meta-clang \
${TOPDIR}/../layers/meta-edgeai \
${TOPDIR}/../layers/meta-tisdk/meta-ti-foundational \
"
(2)
- Apply Device Tree Patch: Apply the required patch to
meta-edgeai/conf/layer.conf to ensure Aquila AM69/TDA4 device trees are included (3):
index 587607cbcd17..b1fa09c93446 100644
--- i/conf/layer.conf
+++ w/conf/layer.conf
@@ -36,6 +36,6 @@ require conf/distro/include/edgeai-source-ipk.inc
TI_TEST_TI_TOOLS:remove:pn-ti-test = "videnc-test-app"
-KERNEL_DEVICETREE_PREFIX:append:j721s2 = "ti/k3-j721s2"
-KERNEL_DEVICETREE_PREFIX:append:j784s4 = "ti/k3-j784s4"
-KERNEL_DEVICETREE_PREFIX:append:j721e = "ti/k3-j721e"
(3)
- Configure local.conf: Add the machine configuration and accept commercial licenses (3):
MACHINE ?= "aquila-am69"
LICENSE_FLAGS_ACCEPTED += "commercial"
(3)
- Build the Edge AI Image: Instead of
tdx-reference-multimedia-image, build the edge AI image (3):
$ bitbake edgeai-reference-image
(3)
U-Boot Configuration Fix
The documentation also mentions that a race condition can occur where remote processors fail to start correctly (1). To resolve this, you need to patch your U-Boot environment to execute boot_rproc after bootcmd_prepare to ensure proper initialization order (1).
Why Your Current Approach Isn’t Working
The issue you’re experiencing occurs because:
- The standard
tdx-reference-multimedia-image doesn’t include the TI Edge AI stack and associated firmware
- The
meta-edgeai-custom layer requires the full TI Edge AI layer stack to function properly
- The
ti-vision-apps recipe is part of the TI Edge AI layers, not the base Toradex BSP
Expected Results
After following this setup, you should have:
- All required firmware files including C71 DSP firmware (
j784s4-c71_0-fw through j784s4-c71_3-fw)
- Proper carveout address alignment between device tree and firmware
- Working remoteproc cores without
bad phdr errors
The edgeai-reference-image will include the complete TI Edge AI stack with TIOVX/vision_apps support and all necessary firmware components (4).
Best regards, Toradex AI Assistant.
Did you like my answer? 
If I answered your question, please tick the “Solution” checkbox.
For follow-up questions to me, please use @ToradexAI in your post.
