zstd is a lot faster when compressing and decompressing, which helps
with iteration speed.
Use the CONVERSIONTYPES machinery builtin rather than re-implementing it
ourselves here.
Signed-off-by: Austin Schuh <austin.linux@gmail.com>
Signed-off-by: Matt Madison <matt@madison.systems>
OE-Core master has udpated how the toolchain virtual providers
are handled, so update the layer accordingly.
This affects the gcc-for-nvcc extra toolchain build we have for
CUDA builds, and a few other recipes that have some explicit
references to the old toolchain virtual/<something> names.
Signed-off-by: Matt Madison <matt@madison.systems>
Starting with the BOARDID and BOARDSKU variables.
For our purposes, they need to be fixed at build time,
and if they don't match the actual target machine, we'll
get difficult-to-diagnose boot failures.
Signed-off-by: Matt Madison <matt@madison.systems>
* Move the mostly-common Orin NX settings to a new orin-nx.inc file
* Set TEGRA_FLASHVAR_CHIP_SKU for machines that differ from the
common default, derived from the chip SKU referenced in the
TEGRA_BUPGEN_SPECS setting
* Convert jetson-agx-orin-devkit-industrial config to using the
agx-orin.inc include file
Signed-off-by: Matt Madison <matt@madison.systems>
This change removes rtl8822 from nvidia-kernel-oot to switch to the
driver and firmware from linux-yocto and linux firmware. It includes
updates to tegra-wifi to pull in the needed packages to the image.
Signed-off-by: Kurt Kiefer <kekiefer@gmail.com>
Custom boards wishing to provide their own devicetree sources will
want to be a provider of virtual/dtb. If this happens, then other
providers of virtual/dtb will not be built. That's a problem, because
the nvidia-kernel-oot recipe also contains device drivers and overlay
devicetree sources that are required.
To solve this, the deployment step of nvidia-kernel-oot is broken
out into a special provider for virtual/dtb. That way, it can easily
be skipped by creating a new virtual/dtb provider and the rest of
nvidia-kernel-oot can still be built and used.
Signed-off-by: Kurt Kiefer <kekiefer@gmail.com>
A number of the out-of-tree modules are named identically to
modules built in-tree, and are intended to replace the in-tree
copies. Mark those with RREPLACES/RCONFLICTS settings when
packaging the modules.
Also add PREFERRED_RPROVIDER settings for those packages,
to point to the out-of-tree copies, to the tegra-common machine
configuration include file.
Signed-off-by: Matt Madison <matt@madison.systems>
* Add a default setting for TEGRA_FLASHVAR_TBCDTB_FILE that resolves
to "@DTBFILE@"
* Update the flash helper to:
- replace @DTBFILE@ with the name of the device tree file in that variable
- remove the TBCDTB-FILE replacement in the storage layout, which isn't needed
- simplify p3701 flashvars handling, similar to what's there for p3767
Signed-off-by: Matt Madison <matt@madison.systems>
The spillover of some boot firmware partitions into
the eMMC was specific to some of the Xavier models,
so we can drop handling of that case.
Signed-off-by: Matt Madison <matt@madison.systems>
Most of the entries in that list have been obsolete since we moved
to using the nvidia-l4t-*.deb packages for installing the tegra-specific
libraries. The only entry we definitely still need is the one for the
gcc-for-nvcc cross-compiler; delete all others for now, and we can
populate more-correct entries when we run into sstate sharing problems.
Signed-off-by: Matt Madison <matt@madison.systems>
This reverts commit 1eec81c883.
With L4T R36.x, we now have a 5.15 kernel as a base, so we no
longer need to do this.
Signed-off-by: Matt Madison <matt@madison.systems>
Drop bup-payload image type from TEGRA_INITRAMFS_FSTYPES, as it's
planned to generate BUP image inside UEFI capsule generation
recipes.
Signed-off-by: Igor Opaniuk <igor.opaniuk@foundries.io>
Starting with R35.5.0, the BUP generator needs the chip SKU passed
to it for generating bootloader updates for Orin modules. Update all
of the Orin configs to reflect that.
Also refactor the AGX Orin configs so we have a separate machine
definition for each of the board SKUs, similar to what's there for
the Orin NX/Nano configs.
Signed-off-by: Matt Madison <matt@madison.systems>
This aligns with the stock L4T/JetPack configuration. Note that
the Android-style separate kernel/kernel-dtb partitions are also
present by default, and will be used by the L4TLauncher UEFI app
to load the kernel and DTB if the extlinux.conf method fails.
Signed-off-by: Matt Madison <matt@madison.systems>
Unlike stock L4T, which has to handle multiple BOARDSKUs for each
module with its prebuilt binary distribution, we can reduce the
number of BUP entries we build per machine, since we have to target
specific BOARDSKUs at build time.
Orin Nano configurations should need to handle only BOARDSKUs 0003
(8GiB) and 0004 (4GiB).
The p3768-0000-p3767-0000 configuration was already trimmed to cover
just the 0000 (16GiB) SKU, so the p3509-a02-p3767-0000 configuration
is updated to match that, since it's just the carrier that's different.
Signed-off-by: Matt Madison <matt@madison.systems>
* Use flash_l4t_t234_nvme.xml as the default layout
* Mark it as having a _rootfs_ab varian
Also, the BUP generation specs for p3768-0000-p3767-0000
are cleaned up to include just the 0000 board SKU, since we
have separate machine configs for the other SKUs.
Signed-off-by: Matt Madison <matt@madison.systems>
R35.5.0 introduces different logic for selecting the PMIC config,
based on board SKU. Update the machine config to expect a sed
rewrite in the flash variable, and update the flash helper to
perform that rewrite.
Signed-off-by: Matt Madison <matt@madison.systems>
As of the nanbield release, OLDEST_KERNEL is set in meta.conf to
version 5.15. This is newer than the version available to us for the
tegra machines in the Jetpack 5 releases. As such, some components
that use this variable (such as glibc) will be build assuming support
for kernel features we don't have.
Override the OLDEST_KERNEL to 5.10, to match the linux-tegra kernel.
Signed-off-by: Kurt Kiefer <kekiefer@gmail.com>
