README: update for L4T R32.3.1/JetPack 4.3

and create a README.md symlink for better formatting
when viewing on Github.

Signed-off-by: Matt Madison <matt@madison.systems>

README

@@ -1,15 +1,17 @@
-OpenEmbedded/Yocto BSP layer for NVIDIA Tegra X1/X2/AGX
+OpenEmbedded/Yocto BSP layer for NVIDIA Jetson TX1/TX2/AGX Xavier/Nano
-=======================================================
+======================================================================
 
 Boards supported:
-* Jetson-TX1 development kit (Linux4Tegra R32.2.3, JetPack 4.2.3)
+* Jetson-TX1 development kit (Linux4Tegra R32.3.1, JetPack 4.3)
-* Jetson-TX2 development kit (Linux4Tegra R32.2.3, JetPack 4.2.3)
+* Jetson-TX2 development kit (Linux4Tegra R32.3.1, JetPack 4.3)
-* Jetson AGX Xavier development kit (Linux4Tegra R32.2.3, JetPack 4.2.3)
+* Jetson AGX Xavier development kit (Linux4Tegra R32.3.1, JetPack 4.3)
-* Jetson Nano development kit (Linux4Tegra R32.2.3, JetPack 4.2.3)
+* Jetson Nano development kit (Linux4Tegra R32.3.1, JetPack 4.3)
+
+Also supported thanks to community support:
+* Jetson-TX2i module (Linux4Tegra R32.3.1, JetPack 4.3)
+* Jetson-TX2 4GB module (Linux4Tegra R32.3.1, JetPack 4.3)
+* Jetson AGX Xavier 8GB module (Linux4Tegra R32.3.1, JetPack 4.3)
 
-Also supported:
-* Jetson-TX2i module (Linux4Tegra R32.2.3, JetPack 4.2.3)
-* Jetson-TX2 4GB module (Linux4Tegra R32.2.3, JetPack 4.2.3)
 
 This layer depends on:
 URI: git://git.openembedded.org/openembedded-core
@@ -30,7 +32,12 @@ PLEASE NOTE
   You must then set the variable NVIDIA_DEVNET_MIRROR to
   "file://path/to/the/downloads" in your build configuration
   (e.g., local.conf) to make them available to your bitbake
-  builds.
+  builds.  This can be the NVIDIA SDK Manager downloads
+  directory, `/home/$USER/Downloads/nvidia/sdkm_downloads`
+
+  **Note** Starting with L4T R32.3 and JetPack 4.3, The Tegra
+  Multimedia API kit has moved to JetPack, so **all builds**
+  now require you to set up an SDK Manager downloads area.
 
 * The SDK Manager downloads a different package of CUDA host-side
   tools depending on whether you are running Ubuntu 16.04
@@ -43,13 +50,20 @@ PLEASE NOTE
   them. Otherwise, the recipes will default to looking for
   the Ubuntu 18.04 package.
 
-* The tensorrt 5.1.6 packages for Xavier are different from
+* The TensorRT 6.0.1 packages for Xavier are different from
   those for TX1/TX2, even though the deb files have the same
-  name.  If you need to build for Xavier and another platform
+  name. To prevent mixups during the build, the recipe here
-  and include tensorrt 5.1.6, create a subdirectory called
+  expects to find the Xavier packages in a `DLA` subdirectory
-  "P2888" under your NVIDIA_DEVNET_MIRROR directory, and copy
+  under `${NVIDIA_DEVNET_MIRROR}`, and non-Xavier packages
-  the Xavier tensorrt packages there. The non-Xavier copies
+  in a `NoDLA` subdirectory.
-  should go in the NVIDIA_DEVNET_MIRROR top level.
+
+  If you need to include TensorRT in your builds, you **must**
+  create the subdirectory and move all of the TensorRT packages
+  downloaded by the SDK Manager there. Xavier example:
+
+      $ cd ~/Downloads/nvidia/sdkm_downloads
+      $ mkdir DLA
+      $ mv tensorrt*.deb libnvinfer*.deb DLA/
 
 * CUDA 10 supports up through gcc 7 only, and some NVIDIA-provided
   binary libraries appear to be compiled with g++ 7 and cause linker

README.md

@@ -0,0 +1 @@
+README