The license checksum changed due to the version, but the license itself
has not changed since the last release (1.1.0)
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
This file was shipped as part of the DDK, but was also supplied by
libwayland so it was deleted here. This file has been removed from
the DDK, so deleting the file manually here is no longer needed.
Signed-off-by: Andrew F. Davis <afd@ti.com>
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
No one in-tree depends on this anymore. This used to be the name of the
package so this was kept here for several years for backwards
compatibility. Remove this from PROVIDES.
Signed-off-by: Andrew F. Davis <afd@ti.com>
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Although not needed for the IMG default EGL implementation this is needed
when building the Mesa EGL version of the DDK.
Signed-off-by: Andrew F. Davis <afd@ti.com>
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Although not needed for the IMG default EGL implementation libexpat
is needed when using the Mesa EGL version.
While here s/libudev/udev as 'libudev' is package shipped as part of the
eudev project, but we use systemd, so we actually depend on 'udev'.
Move the PACKAGE_ARCH and PR defines to a more natural spot that also
lines up with the RGX version of recipe for easier diffs.
One thing we could have done is remove the DEPENDS line. We do not have
any build dependencies as these are pre-made library binaries, but I use
the package rootfs generated by bitbake to build the binaries locally,
so leave it for now.
Signed-off-by: Andrew F. Davis <afd@ti.com>
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
The GBM library is tightly coupled with the system graphics systems,
other hardware GPU vendors (including our Rogue DDK) and Mesa all deploy
a version of the library as part of their GPU driver. We SGX does the
same. This should ease integration for us as we now don't rely on a
separate project for this library and our deployment strategy can match
across Mesa/RGX recipes, allowing for better re-use of common components.
Signed-off-by: Andrew F. Davis <afd@ti.com>
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
* Configure the linux-demo to use a ramfs for the linux inmate.
* Using JH_RAMFS_IMAGE, this can be set in some global conf.
Signed-off-by: Jacob Stiffler <j-stiffler@ti.com>
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Create a package containing the jailhouse tools which are used to work
with the cells and inmates. This recipe provides a good example for
using the jailhouse-cell-linux tool.
The tools require the pyjailhouse python modules, so package them as
well.
Signed-off-by: Jacob Stiffler <j-stiffler@ti.com>
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Correcting the corelist. There is no mpu2_x or mpu3_x instead they
should be mcu2_x/mcu3_x which is present in j7-evm
Signed-off-by: Mahesh Radhakrishnan <m-radhakrishnan2@ti.com>
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
There is no mpu2_1,mpu3_0,mpu3_1 for j7-evm. They should have been
mcu2_1,mcu3_0 and mcu3_1 respectively.
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
The default TI_PDK_LIMIT_CORES set in conf/machine/includes/am65xx.inc
doesn't contain mcu1_1 which sciclient supports for am65x. Hence
appending it in the recipe.
Signed-off-by: Mahesh Radhakrishnan <m-radhakrishnan2@ti.com>
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
* extend to am65xx family
* and correct am65xx cell and inmate filenames.
* remove earlycon bootargs for am65xx
Signed-off-by: Jacob Stiffler <j-stiffler@ti.com>
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
