- Enable bootscript templating using `envsubst`.
- Add generic bootscript.
- Fix: `test` does not support `==`, but deceptively returns `true` in case you use it.
- De-duplicate kernel commandline console options.
- Change debug console from ttyS2 to ttyS0
- Remove SERIALCON from board configs
- Set console based on BOOT_SOC
- Enable UART0 in device tree for nanopi-r76s
* cleanup comments, linebreaks...
* two more comments
* `BOOTFS_TYPE` has never been set as far as I can see, therefore no need to unset.
* some more comments...
* comment, simplify logic
* even more comments
* just a few more
* comments, description for elish
* merge uboot functions. @SuperKali review
Instead of using the TI k3.conf and overriding configuration, add a new
config for BeagleBoard.org Linux and U-Boot and use it for the boards
that use that Linux/U-Boot.
Signed-off-by: Andrew Davis <afd@ti.com>
Vendors using TI K3 devices may use kernels downstream from the TI vendor
kernel and U-Boot. To make this easier factor out from k3.conf all other
configuration that would be common to any kernel and U-Boot used.
Signed-off-by: Andrew Davis <afd@ti.com>
The latest Beagle kernel has some support now for the BeagleY-AI, update
the current branch to this version. This also allows us to re-use the
default linux-k3-beagle-current.conf as our LINUXCONFIG. This was set
to re-use the TI kernel before which was incorrect.
As BOOTPATCHDIR is based on LINUXFAMILY we should also keep our Beagle
U-Boot versions in sync, the latest Beagle U-Boot for pocketbeagle2
supports BeagleY-AI also, so switch to that.
Signed-off-by: Andrew Davis <afd@ti.com>
The "current" and "edge" branches point to the same kernel and u-boot.
No need to have both so just remove "edge" as it was not any more
up-to-date than the standard "current" branch.
Signed-off-by: Andrew Davis <afd@ti.com>
- `=m` is not a valid value for options that are `bool`
- so change them to `=y` manually
- from olddefconfig:
- `.config:82:warning: symbol value 'm' invalid for BT_RFCOMM_TTY`
- `.config:84:warning: symbol value 'm' invalid for BT_BNEP_MC_FILTER`
- `.config:85:warning: symbol value 'm' invalid for BT_BNEP_PROTO_FILTER`
- `.config:87:warning: symbol value 'm' invalid for BT_HS`
- `.config:89:warning: symbol value 'm' invalid for BT_LEDS`
- `.config:90:warning: symbol value 'm' invalid for BT_MSFTEXT`
- `.config:92:warning: symbol value 'm' invalid for BT_HCIBTUSB_AUTOSUSPEND`
- `.config:93:warning: symbol value 'm' invalid for BT_HCIBTUSB_MTK`
- `.config:96:warning: symbol value 'm' invalid for BT_HCIUART_BCSP`
- `.config:97:warning: symbol value 'm' invalid for BT_HCIUART_ATH3K`
- `.config:99:warning: symbol value 'm' invalid for BT_HCIUART_BCM`
- `.config💯warning: symbol value 'm' invalid for BT_HCIUART_RTL`
- `.config:101:warning: symbol value 'm' invalid for BT_HCIUART_QCA`
- `.config:102:warning: symbol value 'm' invalid for BT_HCIUART_AG6XX`
- `.config:250:warning: symbol value 'm' invalid for AIC_WLAN_SUPPORT`
- this is to workaround e3ec97c3ab
- I talked to the maintainer (Wei Liu) upstream, they are aware of the problem, but haven't a solution
- the cloud variants are not affected (already had =y)
eMMC installs on the A20-OLinuXino-LIME2 eMMC hang during boot.
This is because we need the eMMC power sequencing module in early boot.
Compile it in to the kernel so we can boot from eMMC again.