This commit fixes the following problems with the RPU device tree:
1. The core type of the RPU of ZynqMP SoC is Cortex-R5F, not
Cortex-R4.
2. RPU and APU use different interrupt controllers (PL390 GICv1 and
GIC-400 GICv2, respectively) mapped to the same CPU local bus address
region but with different offsets for the distributor and CPU
interrupt control register sets. The GIC address mapping specified by
the current dts is that of an APU and does not apply to the PL390
GICv1 of an RPU (refer to the "Zynq UltraScale+ Devices Register
Reference" document from Xilinx for more information).
For more details, refer to the issue #20217.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
ZynqMP SoC embeds two separate processor types: Cortex-R for RPU and
Cortex-A for APU.
Since the current Zephyr architecture cannot support AMP of Cortex-R
and Cortex-A within one project, the RPU and APU should be considered
separate platforms.
This commit relocates the device tree nodes that are not common between
RPU and APU to a separate dtsi file (zynqmp_rpu.dtsi).
When Cortex-A53 APU support is added in the future, an additional dtsi
file (zynqmp_apu.dtsi) for specifying the APU device tree should be
added.
For more details, refer to the issue #20217.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
The GIC-400 driver currently only supports SPIs because the (32) offset
for the INTIDs is hard-coded in the driver. At the driver level there is
no really difference between PPIs and SPIs so we can easily extend the
driver to support PPIs as well.
This is useful if we want to add support for the ARM Generic Timers that
use INTIDs in the PPI range.
SPI interrupts are in the range [0-987]. PPI interrupts are in the range
[0-15].
This commit adds interrupt 'type' cell to the GIC device tree binding
and changes the 'irq' cell to use interrupt type-specific index, rather
than a linear IRQ number.
The 'type'+'irq (index)' combo is automatically fixed up into a linear
IRQ number by the scripts/dts/gen_defines.py script.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
1. Replace the non-existent CPU device binding ("Cortex-R") specified
by the CPU node with a proper one.
2. Relocate CPU node declaration to SoC dtsi:
The CPU node should be declared in the SoC dtsi because the core
type is SoC-dependent. In fact, this is exactly how it is done in
the Cortex-M port.
3. Remove core_intc (supposedly Cortex-R VIC):
Unlike the NVIC of Cortex-M, the VIC of Cortex-R is not a true
interrupt controller in the conventional sense and merely acts as
a CPU input port for aggregated interrupt request and vector index
signals. For this reason, there is no point in declaring it in the
device tree and specifying it as an interrupt parent. All SoCs
incorporating Cortex-R implement a separate true interrupt
controller (for instance, GIC for Zynq MPSoC and VIM for Hercules).
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
'interrupt-parent' should contain just the phandle of the node
interrupts are sent to.
This node (gic: interrupt-controller@f9010000) doesn't generate any
interrupts, so the 'interrupt-parent' value is never used (this is why
it wasn't caught). It'll give an error later with 'interrupt-parent'
declared as 'type: phandle' in bindings though.
Don't know what was intended. Just remove the 0.
Signed-off-by: Ulf Magnusson <Ulf.Magnusson@nordicsemi.no>
This commit adds support for the Zynq UltraScale+ MPSoC as a qemu based
platform for Cortex-R based testing. This SoC only supports an
interrupt controller and serial port for limited testing.
Signed-off-by: Bradley Bolen <bbolen@lexmark.com>
Stephanos Ioannidis