feat(arm_fpga): write UART baud base clock frequency into DTB

Since we now autodetect the actual system frequency, which is also used
as the base for the UART baudrate generation, we should update t

feat(arm_fpga): write UART baud base clock frequency into DTB

Since we now autodetect the actual system frequency, which is also used
as the base for the UART baudrate generation, we should update the value
currently hard-coded in the DT. Otherwise Linux will reprogram the
divider using a potentially wrong base rate, which breaks the UART
output.

Find the DT node referenced by the UART node as the clock rate, and set
the "clock-frequency" property in that node to the detected system
frequency. This will let Linux reprogram the divider to the same value,
preserving the actual baudrate.

Change-Id: Ib5a936849f2198577b86509f032751d5386ed2f8
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

feat(arm_fpga): query PL011 to learn system frequency

The Arm FPGAs run in mostly one clock domain, which is used for the CPU
cores, the generic timer, and also the UART baudrate base clock. This
si

feat(arm_fpga): query PL011 to learn system frequency

The Arm FPGAs run in mostly one clock domain, which is used for the CPU
cores, the generic timer, and also the UART baudrate base clock. This
single clock can have different rates, to compensate for different IP
complexity. So far most images used 10 MHz, but different rates start to
appear.

To avoid patching both the arch timer frequency and UART baud base fixed
clock in the DTB manually, we would like to set the clock rate
automatically. Fortunately the SCP firmware has the actual clock rate
hard coded, and already programs the PL011 UART baud divider register
with the correct value to achieve a 38400 bps baudrate.

So read the two PL011 baudrate divider values and re-calculate the
original base clock from there, to use as the arch timer frequency. If
the arch timer DT node contains a clock-frequency property, we use that
instead, to support overriding and disabling this autodetection.

Change-Id: I9857fbb418deb4644aeb2816f1102796f9bfd3bb
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

refactor(arm_fpga): move command line code into separate function

The code dealing with finding the command line and inserting that into
the DTB is somewhat large, and drowns the other DT handlers i

refactor(arm_fpga): move command line code into separate function

The code dealing with finding the command line and inserting that into
the DTB is somewhat large, and drowns the other DT handlers in our
fpga_prepare_dtb() function.

Move that code into a separate function, to improve readability.

Change-Id: I828203c4bb248d38a2562fcb6afdefedf3179f8d
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

feat(arm_fpga): add ITS autodetection

Some FPGAs come with a GIC that has an ITS block configured. Since the
ITS sits between the distributor and redistributors, we can autodetect
that, and already

feat(arm_fpga): add ITS autodetection

Some FPGAs come with a GIC that has an ITS block configured. Since the
ITS sits between the distributor and redistributors, we can autodetect
that, and already adjust the GICR base address.

To also make this ITS usable, add an ITS node to our base DTB, and
remove that should we not find an ITS during the scan for the
redistributor. This allows to use the same TF-A binary for FPGA images
with or without an ITS.

Change-Id: I4c0417dec7bccdbad8cbca26fa2634950fc50a66
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

feat(arm_fpga): determine GICR base by probing

When an Arm Ltd GIC (Arm GIC-[567]00) is instantiated with one or more
ITSes, the ITS MMIO frames appear between the distributor and
redistributor addr

feat(arm_fpga): determine GICR base by probing

When an Arm Ltd GIC (Arm GIC-[567]00) is instantiated with one or more
ITSes, the ITS MMIO frames appear between the distributor and
redistributor addresses. This makes the beginning of the redistributor
region dependent on the existence and number of ITSes.

To support various FPGA images, with and without ITSes, probe the
addresses in question, to learn whether they accommodate an ITS or a
redistributor. This can be safely done by looking at the PIDR[01]
registers, which contain an ID code for each region, documented in the
Arm GIC TRMs.

We try to find all ITSes instantiated, and skip either two or four 64K
frames, depending on GICv4.1 support. At some point we will find the
first redistributor; this address we then update in the DTB.

Change-Id: Iefb88c2afa989e044fe0b36b7020b56538c60b07
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

feat(libfdt): also allow changing base address

For platforms where we don't know the number of cores at compile time,
the size of the GIC redistributor frame is then also undetermined, since
it depe

feat(libfdt): also allow changing base address

For platforms where we don't know the number of cores at compile time,
the size of the GIC redistributor frame is then also undetermined, since
it depends on this number of cores.
On top of this the GICR base address can also change, when an unknown
number of ITS frames (including zero) take up space between the
distributor and redistributor.

So while those two adjustments are done for independent reasons, the
code for doing so is very similar, so we should utilise the existing
fdt_adjust_gic_redist() function.

Add an (optional) gicr_base parameters to the prototype, so callers can
choose to also adjust this base address later, if needed.

Change-Id: Id39c0ba83e7401fdff1944e86950bb7121f210e8
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

fix(arm_fpga): avoid re-linking from executable ELF file

When we build the convenience firmware package file for the Arm FPGA
boards (bl31.axf), we combine trampolines, the DTB and the actual BL31
c

fix(arm_fpga): avoid re-linking from executable ELF file

When we build the convenience firmware package file for the Arm FPGA
boards (bl31.axf), we combine trampolines, the DTB and the actual BL31
code into one ELF file, which is more a "container with load addresses"
than an actual executable. So far ld was fine with us using bl31.elf as
an input file, but binutils 2.35 changed that and complains about
taking an *executable* ELF file as in *input* to the linker:
-----------------
aarch64-none-elf-ld.bfd: cannot use executable file 'build/arm_fpga/debug/./bl31/bl31.elf' as input to a link
-----------------

Fortunately we don't need the actual BL31 ELF file for *that* part of
the linking, so can use the just created bl31.bin binary version of it.
Actually that shrinks the file, as we needlessly included the .BSS
section in the final file before.

Using the binary works with both older and newer toolchains versions, so
let's do this unconditionally.

Change-Id: Ib7e697f8363499123f7cb860f118f182d0830768
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

build(fdt-wrappers): introduce FDT wrappers makefile

This has been introduced to simplify dependencies on the FDT wrappers.
We generally want to avoid pulling in components on a file-by-file
basis,

build(fdt-wrappers): introduce FDT wrappers makefile

This has been introduced to simplify dependencies on the FDT wrappers.
We generally want to avoid pulling in components on a file-by-file
basis, particularly as we are trying to draw conceptual boxes around
components in preparation for transitioning the build system to CMake,
where dependencies are modelled on libraries rather than files.

Signed-off-by: Chris Kay <chris.kay@arm.com>
Change-Id: Idb7ee05a9b54a8caa3e07f36e608867e20b6dcd5

show more ...

fix(arm_fgpa): allow build after MAKE_* changes

Commit 434d0491c550 ("refactor(makefile): remove BL prefixes in build
macros") changed the MAKE_S macro to expect "bl31" instead of just "31".

Adjust

fix(arm_fgpa): allow build after MAKE_* changes

Commit 434d0491c550 ("refactor(makefile): remove BL prefixes in build
macros") changed the MAKE_S macro to expect "bl31" instead of just "31".

Adjust our calls to MAKE_S and MAKE_LD to fix the build for arm_fpga.

Change-Id: I2743e421c10eaecb39bfa4515ea049a1b8d18fcb
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

Merge changes from topic "arm_fpga_resmem" into integration

* changes:
fix(arm_fpga): reserve BL31 memory
fix(arm_fpga): limit BL31 memory usage

fix(arm_fpga): streamline generated axf file

For convenience we let the build system generate an ELF file (named
bl31.axf), containing all the trampolines, BL31 code and the DTB in one
file. This ca

fix(arm_fpga): streamline generated axf file

For convenience we let the build system generate an ELF file (named
bl31.axf), containing all the trampolines, BL31 code and the DTB in one
file. This can be fed directly into the FPGA payload tool, and it will
load the bits at the right addresses.
Since this ELF file is more used as a "container with load addresses",
there is no need for normal ELF features like alignment or a symbol
table.

Remove unnecessary sections from that output file, by doing a static
"link", dropping the NOBITS stacks section, and by adding "-n" to the
linker command line (to avoid page alignment). This trims the generated
.axf file, and makes it smaller.

Change-Id: I5768543101d667fb4a3b70e60b08cfe970d2a2b6
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

feat(arm_fpga): add kernel trampoline

The arm64 Linux kernel needed to be loaded at a certain offset within any
2MB aligned region; this value was configured at compile time and stored
in the Linux

feat(arm_fpga): add kernel trampoline

The arm64 Linux kernel needed to be loaded at a certain offset within any
2MB aligned region; this value was configured at compile time and stored
in the Linux kernel image header. The default value was always 512KiB,
so this is the value we use in the TF-A build system for the kernel
load address.
However the whole scheme around the TEXT_OFFSET changed in Linux v5.8:
Linux kernels became fully relocatable, so this value is largely ignored
now, and its default value changed to 0. The only remainder is a warning
message at boot time in case there is a mismatch:
[Firmware Bug]: Kernel image misaligned at boot, please fix your bootloader!

To avoid this warning, and to make TF-A Linux kernel boot protocol
compliant, we should load newer kernels to offset 0 of a 2 MB
region. This can be done by the user at FPGA boot time, but BL31 needs
to know about this address. As we can't change the build default to 0
without breaking older kernels, we should try to make a build dealing
with both versions:

This patch introduces a small trampoline code, which gets loaded at
512KB of DRAM, and branches up to 2MB. If users load their newer
kernels at 2MB, this trampoline will cover them. In case an older kernel
is loaded at 512KB, it will overwrite this trampoline code, so it would
still work.

Change-Id: If49ca86f5dca380036caf2555349748722901277
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

fix(arm_fpga): reserve BL31 memory

Embarrassingly we never told the non-secure world that secure firmware
lives in the first few hundred KBs of DRAM, so any non-secure payload
could happily overwrit

fix(arm_fpga): reserve BL31 memory

Embarrassingly we never told the non-secure world that secure firmware
lives in the first few hundred KBs of DRAM, so any non-secure payload
could happily overwrite TF-A, and we couldn't even blame it.

Advertise the BL31 region in the reserved-memory DT node, so non-secure
world stays out of it.

This fixes Linux booting on FPGAs with less memory than usual.

Change-Id: I7fbe7d42c0b251c0ccc43d7c50ca902013d152ec
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

fix(arm_fpga): limit BL31 memory usage

At the moment we specified the BL31 memory limits to 1MB; since we
typically have gigabytes of DRAM, we can be quite generous.

However the default parameters

fix(arm_fpga): limit BL31 memory usage

At the moment we specified the BL31 memory limits to 1MB; since we
typically have gigabytes of DRAM, we can be quite generous.

However the default parameters expect the devicetree binary at
0x80070000, so we should actually make sure we have no code or data
beyond that point.

Limit the ARM FPGA BL31 memory footprint to this available 7*64K region.
We stay within the limit at the moment, with more than half of it
reserved for stacks, so this could be downsized later should we run
into problems.

The PIE addresses stay as they are, since the default addresses do not
apply there anywhere, and the build is broken anyway.

Change-Id: I7768af1a93ff67096f4359fc5f5feb66464bafaa
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

feat(arm_fpga): support GICv4 images

Up until now we relied on the GICs used in our FPGA images to be GICv3
compliant, without the "direct virtual injection" feature (aka GICv4)
enabled.
To support

feat(arm_fpga): support GICv4 images

Up until now we relied on the GICs used in our FPGA images to be GICv3
compliant, without the "direct virtual injection" feature (aka GICv4)
enabled.
To support newer images which have GICv4 compliant GICs, enable the
newly introduced GICv4 detection code, and use that also when we adjust
the redistributor region size in the devicetree.

This allows the same BL31 image to be used with GICv3 or GICv4 FPGA
images.

Change-Id: I9f6435a6d5150983625efe3650a8b7d1ef11b1d1
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

feat(gicv3): detect GICv4 feature at runtime

At the moment we have a GIC_ENABLE_V4_EXTN build time variable to
determine whether the GIC interrupt controller is compliant to version
4.0 of the spec

feat(gicv3): detect GICv4 feature at runtime

At the moment we have a GIC_ENABLE_V4_EXTN build time variable to
determine whether the GIC interrupt controller is compliant to version
4.0 of the spec or not. This just changes the number of 64K MMIO pages
we expect per redistributor.

To support firmware builds which run on variable systems (emulators,
fast model or FPGAs), let's make this decision at runtime.
The GIC specification provides several architected flags to learn the
size of the MMIO frame per redistributor, we use GICR_TYPER[VLPI] here.

Provide a (static inline) function to return the size of each
redistributor.
We keep the GIC_ENABLE_V4_EXTN build time variable around, but change
its meaning to enable this autodetection code. Systems not defining this
rely on a "pure" GICv3 (as before), but platforms setting it to "1" can
now deal with both configurations.

Change-Id: I9ede4acf058846157a0a9e2ef6103bf07c7655d9
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

fix(plat/arm_fpga): enable AMU extension

As done recently for plat/tc0 in b5863cab9, enable AMU explicitly.
This is necessary as the recent changes that enable SVE for the secure
world disable AMU b

fix(plat/arm_fpga): enable AMU extension

As done recently for plat/tc0 in b5863cab9, enable AMU explicitly.
This is necessary as the recent changes that enable SVE for the secure
world disable AMU by default in the CPTR_EL3 reset value.

Change-Id: Ie3abf1dee8a4e1c8c39f934da8e32d67891f5f09
Signed-off-by: Tom Cosgrove <tom.cosgrove@arm.com>

show more ...

fix: rename Matterhorn, Matterhorn ELP, and Klein CPUs

This patch renames the Matterhorn, Matterhorn ELP, and Klein CPUs to
Cortex A710, Cortex X2, and Cortex A510 respectively.

Signed-off-by: John

fix: rename Matterhorn, Matterhorn ELP, and Klein CPUs

This patch renames the Matterhorn, Matterhorn ELP, and Klein CPUs to
Cortex A710, Cortex X2, and Cortex A510 respectively.

Signed-off-by: John Powell <john.powell@arm.com>
Change-Id: I056d3114210db71c2840a24562b51caf2546e195

show more ...

Add "_arm" suffix to Makalu ELP CPU lib

ELP processors can sometimes have different MIDR values or features so
we are adding the "_arm" suffix to differentiate the reference
implementation from othe

Add "_arm" suffix to Makalu ELP CPU lib

ELP processors can sometimes have different MIDR values or features so
we are adding the "_arm" suffix to differentiate the reference
implementation from other future versions.

Signed-off-by: John Powell <john.powell@arm.com>
Change-Id: Ieea444288587c7c18a397d279ee4b22b7ad79e20

show more ...

Add Cortex_A78C CPU lib

Add basic support for Cortex_A78C CPU.

Signed-off-by: Bipin Ravi <bipin.ravi@arm.com>
Change-Id: Id9e41cbe0580a68c6412d194a5ee67940e8dae56

Add Makalu ELP CPU lib

Add basic support for Makalu ELP processor core.

Signed-off-by: John Powell <john.powell@arm.com>
Change-Id: I7b1ddbb8dd43326ecb8ff188f6f8fcf239826a93

Add Makalu CPU lib

Add basic support for Makalu CPU.

Signed-off-by: John Powell <john.powell@arm.com>
Change-Id: I4e85d425eedea499adf585eb8ab548931185043d

Add support for Neoverse-N2 CPUs.

Enable basic support for Neoverse-N2 CPUs.

Signed-off-by: Javier Almansa Sobrino <javier.almansasobrino@arm.com>
Change-Id: I498adc2d9fc61ac6e1af8ece131039410872e8

Add support for Neoverse-N2 CPUs.

Enable basic support for Neoverse-N2 CPUs.

Signed-off-by: Javier Almansa Sobrino <javier.almansasobrino@arm.com>
Change-Id: I498adc2d9fc61ac6e1af8ece131039410872e8ad

show more ...

Rename Neoverse Zeus to Neoverse V1

Change-Id: Ieb411e2f8092fa82062e619305b680673a8f184f
Signed-off-by: Jimmy Brisson <jimmy.brisson@arm.com>

Rename Cortex Hercules AE to Cortex 78 AE

Change-Id: Ic0ca51a855660509264ff0d084c068e1421ad09a
Signed-off-by: Jimmy Brisson <jimmy.brisson@arm.com>