Andre Przywara [Wed, 16 Sep 2020 16:13:33 +0000 (17:13 +0100)]
arm_fpga: Add post-build linker script
For the Arm Ltd. FPGAs to run, we need to load several payloads into the
FPGA's memory:
- Some trampoline code at address 0x0, to jump to BL31's entry point.
- The actual BL31 binary at the beginning of DRAM.
- The (generic) DTB image to describe the hardware.
- The actual non-secure payloads (kernel, ramdisks, ...)
The latter is application specific, but the first three blobs are rather
generic.
Since the uploader tool supports ELF binaries, it seems helpful to
combine these three images into one .axf file, as this also simplifies
the command line.
Add a post-build linker script, that combines those three bits into one
ELF file, together with their specific load addresses.
Include a call to "ld" with this linker script in the platform Makefile,
so it will be build automatically. The result will be called "bl31.axf".
Change-Id: I4a90da16fa1e0e83b51d19e5b1daf61f5a0bbfca Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Andre Przywara [Mon, 3 Aug 2020 12:06:38 +0000 (13:06 +0100)]
arm_fpga: Add ROM trampoline
The application cores of the FPGAs used in Arm Ltd. start execution at
address 0x0. This is the location of some (emulated) ROM area (which can
be written to by the uploading tool).
Since the arm_fpga port is configured to run from DRAM, we load BL31 to
the beginning of DRAM (mapped at 2GB). This requires some small
trampoline code in the "ROM" to jump to the BL31 entry point.
To avoid some extra magic binary, add a tiny assembly file with that
trivial jump instruction to the tree, so this binary can be created
alongside BL31.
Change-Id: I9e4439fc0f093fa24dd49a8377c9edb030fbb477 Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Andre Przywara [Mon, 3 Aug 2020 11:54:58 +0000 (12:54 +0100)]
arm_fpga: Add devicetree file
The FPGA images used in Arm Ltd. focus on CPU cores, so they share a
common platform, with a minimal set of peripherals (interconnect, GIC,
UART).
This allows to support most platforms with a single devicetree file.
The topology and number of CPU cores differ, but those will added at
runtime, in BL31. Other adjustments (GICR size, SPE node, command line)
are also done at this point.
Add the common devicetree file to TF-A's build system, so it can be
build together with BL31. At runtime, the resulting .dtb file should be
uploaded to the address given with FPGA_PRELOADED_DTB_BASE at build time.
Change-Id: I3206d6131059502ec96896e95329865452c9d83e Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Andre Przywara [Mon, 3 Aug 2020 11:55:28 +0000 (12:55 +0100)]
arm_fpga: Remove SPE PMU DT node if SPE is not available
The Statistical Profiling Extension (SPE) is an architectural feature we
can safely detect at runtime. However it still relies on one piece of
platform-specific information: the interrupt line it is connected
to. This requires SPE to be described in a devicetree node.
Since SPE support varies with the CPU cores found on an FPGA image, we
should detect the presence of SPE at runtime, and remove a potentially
existing SPE PMU node from the DT.
This allows to always have the SPE node in a generic devicetree file,
without risking exposing it on a CPU without this feature.
Change-Id: I73d83ea8509b03fe7bba20b9cce8d1335035fa31 Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Andre Przywara [Mon, 24 Aug 2020 17:34:50 +0000 (18:34 +0100)]
arm_fpga: Adjust GICR size in DT to match number of cores
The size of a GICv3 redistributor region depends on the number of
cores in the system. For the ARM FPGA port, we detect the topology at
runtime, and adjust the CPU DT nodes accordingly.
Now the size of the GICR region must also be adjusted, or Linux will
fail to initialise the GICv3.
Use the newly introduced function to overwrite the GICR size entry in
the GICv3 reg property. We count the number of existing cores by
iterating over the GICR frames until we find the LAST bit set in TYPER.
Change-Id: Ib69565600859de9b1b15ceb8495172cd26d16fce Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Andre Przywara [Mon, 24 Aug 2020 17:28:44 +0000 (18:28 +0100)]
fdt: Add function to adjust GICv3 redistributor size
We now have code to detect the CPU topology at runtime, and can also
populate the CPU nodes in a devicetree accordingly. This is used by the
ARM FPGA port, for instance.
But also a GICv3 compatible interrupt controller provides MMIO frames
per core, so the size of this region needs to be adjusted in the DT,
to match the number of cores as well.
Provide a generic function to find the GICv3 interrupt controller in
the DT, then adjust the "reg" entry to match the number of detected
cores. Since the size of the GICR frame per cores differs between
GICv4 and GICv3, this size is supplied as a parameter to the function.
The caller should determine the applicable value by either hardcoding
it or by observing GICR_TYPER.VLPIS.
Change-Id: Ic2a6445c2c5381a36bf24263f52fcbefad378c05 Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Andre Przywara [Mon, 7 Sep 2020 13:53:58 +0000 (14:53 +0100)]
drivers: arm: gicv3: Allow detecting number of cores
A GICv3 interrupt controller will be instantiated for a certain number
of cores. This will result in the respective number of GICR frames. The
last frame will have the "Last" bit set in its GICR_TYPER register.
For platforms with a topology unknown at build time (the Arm FPGAs, for
instance), we need to learn the number of used cores at runtime, to size
the GICR region in the devicetree accordingly.
Add a generic function that iterates over all GICR frames until it
encounters one with the "Last" bit set. It returns the number of cores
the GICv3 has been configured for.
Change-Id: I79f033c50dfc1c275aba7122725868811abcc4f8 Signed-off-by: Andre Przywara <andre.przywara@arm.com>
MCKPROT hardening in RCC mandates that both bits RCC[TZEN] and
RCC[MCKPROT] are enabled. This change fixes stm32mp1_rcc_is_mckprot()
to check both bits, not RCC[MCKPROT] only.
This change also updates stm32mp1_rcc_is_secure() for consistency.
Yann Gautier [Fri, 12 Jun 2020 12:14:26 +0000 (14:14 +0200)]
mmc: st: clear some flags before sending a command
The ICR static flags are cleared before sending a command.
The SDMMC_DCTRLR register is set to 0 if no data is expected on a given
command or on the next command in case of CMD55.
Lionel Debieve [Wed, 26 Aug 2020 14:17:02 +0000 (16:17 +0200)]
nand: raw_nand: fix timeout issue in nand_wait_ready
nand_wait_ready is called with a millisecond delay
but the timeout used a micro second. Fixing the conversion
in the timeout call.
The prototype of the function is also changed to use an unsigned int
parameter.
Select the Log Level for the Event Log Dump on Measured Boot at build time.
Builds in Debug mode with Measured Boot enabled might run out of trusted
SRAM. This patch allows to change the Log Level at which the Measured Boot
driver will dump the event log, so the latter can be accessed even on
Release builds if necessary, saving space on RAM.
Signed-off-by: Javier Almansa Sobrino <javier.almansasobrino@arm.com>
Change-Id: I133689e313776cb3f231b774c26cbca4760fa120
SPMC: adjust the number of EC context to max number of PEs
According to [1] and in context of FF-A v1.0 a secure partition must
have either one EC (migratable UP) or a number of ECs equal to the
number of PEs (pinned MP). Adjust the SPMC manifest such that the
number of ECs is equal to the number of PEs.
Implemented a parser which populates the properties of
the CoT descriptors as per the binding document [1].
'COT_DESC_IN_DTB' build option is disabled by default and can
be enabled in future for all Arm platforms by making necessary
changes in the memory map.
Currently, this parser is tested only for FVP platform.
Andre Przywara [Fri, 11 Sep 2020 08:18:09 +0000 (09:18 +0100)]
SPE: Fix feature detection
Currently the feature test for the SPE extension requires the feature
bits in the ID_AA64DFR0 register to read exactly 0b0001.
However the architecture guarantees that any values greater than 0
indicate the presence of a feature, which is what we are after in
our spe_supported() function.
Change the comparison to include all values greater than 0.
This fixes SPE support in non-secure world on implementations which
include the Scalable Vector Extension (SVE), for instance on Zeus cores.
Change-Id: If6cbd1b72d6abb8a303e2c0a7839d508f071cdbe Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Sami Mujawar [Thu, 30 Apr 2020 11:40:22 +0000 (12:40 +0100)]
Fix fiptool packaging issue on windows
Windows does not have a standard getopt implementation. To address
this an equivalent implementation has been provided in win_posix.c
However, the implementation has an issue with option processing as
described below.
Long option names may be abbreviated if the abbreviation is unique
or an exact match for some defined option.
Since some options can be substring of other options e.g. "scp-fw"
option is a substring of "scp-fwu-cfg", we need to identify if an
option is abbreviated and also check for uniqueness. Otherwise if
a user passes --scp-fw as an option, the "scp-fwu-cfg" option may
get selected, resulting in an incorrectly packaged FIP.
This issue has been be fixed by:
- First searching for an exact match.
- If exact match was not found search for a abbreviated match.
By doing this an incorrect option selection can be avoided.
Change-Id: I22f4e7a683f3df857f5b6f0783bf9b03a64a0bcc Signed-off-by: Sami Mujawar <sami.mujawar@arm.com>
Neoverse N1 erratum 1868343 is a Cat B erratum, present in older
revisions of the Neoverse N1 processor core. The workaround is to
set a bit in the CPUACTLR_EL1 system register, which delays instruction
fetch after branch misprediction. This workaround will have a small
impact on performance.
Ruari Phipps [Mon, 10 Aug 2020 14:53:45 +0000 (15:53 +0100)]
SPM: Get rid of uint32_t array representation of UUID
UUID's in the device tree files were stored in little endian. So
to keep all entries in these files RFC 4122 compliant, store them in
big endian then convert it to little endian when they are read so they
can be used in the UUID data structure.
Max Shvetsov [Tue, 25 Aug 2020 10:50:18 +0000 (11:50 +0100)]
Fix: fixing coverity issue for SPM Core.
spmd_get_context_by_mpidr was using potentially negative value as an
array index. plat_core_pos_by_mpidr could return -1 on failure which is
utilized by some platforms.
Signed-off-by: Max Shvetsov <maksims.svecovs@arm.com>
Change-Id: I7f8827e77f18da389c9cafdc1fc841aba9f03120
Issue seems to be occurred when 'ROT key' is generated before creating
the platform build folder(for e.g.build/qemu/debug).
Changes are made to fix this issue by adding orderly dependancy of
the platform folder for the 'ROT key' creation which ensures that
platform folder is created before generating 'ROT key'.
Signed-off-by: Manish V Badarkhe <Manish.Badarkhe@arm.com>
Change-Id: I20c82172dde84e4c7f2373c0bd095d353f845d38
This patch: fafd3ec9c assumes that tools must build from
the main makefile folder.
This assumption leads to the error when somebody wants to
build a tool from the tool's folder.
Hence changes are done to provide the default binary name
in the tool's makefile.
Change-Id: Iae570a7f8d322151376b6feb19e739300eecc3fc Signed-off-by: Manish V Badarkhe <Manish.Badarkhe@arm.com>
Andre Przywara [Thu, 3 Sep 2020 10:04:39 +0000 (11:04 +0100)]
libc: memset: improve performance by avoiding single byte writes
Currently our memset() implementation is safe, but slow. The main reason
for that seems to be the single byte writes that it issues, which can
show horrible performance, depending on the implementation of the
load/store subsystem.
Improve the algorithm by trying to issue 64-bit writes. As this only
works with aligned pointers, have a head and a tail section which
covers unaligned pointers, and leave the bulk of the work to the middle
section that does use 64-bit writes.
Put through some unit tests, which exercise all combinations of nasty
input parameters (pointers with various alignments, various odd and even
sizes, corner cases of content to write (-1, 256)).
Change-Id: I28ddd3d388cc4989030f1a70447581985368d5bb Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Sandeep Tripathy [Mon, 17 Aug 2020 14:52:13 +0000 (20:22 +0530)]
psci: utility api to invoke stop for other cores
The API can be used to invoke a 'stop_func' callback for all
other cores from any initiating core. Optionally it can also
wait for other cores to power down. There may be various use
of such API by platform. Ex: Platform may use this to power
down all other cores from a crashed core.
Trace analysis of FVP_Base_AEMv8A 0.0/6063 model
running in Aarch32 mode with the build options
listed below:
TRUSTED_BOARD_BOOT=1 GENERATE_COT=1
ARM_ROTPK_LOCATION=devel_ecdsa KEY_ALG=ecdsa
ROT_KEY=plat/arm/board/common/rotpk/arm_rotprivk_ecdsa.pem
shows that when auth_signature() gets called
71.99% of CPU execution time is spent in memset() function
written in C using single byte write operations,
see lib\libc\memset.c.
This patch introduces new libc_asm.mk makefile which
replaces C memset() implementation with assembler
version giving the following results:
- for Aarch32 in auth_signature() call memset() CPU time
reduced to 20.56%.
The number of CPU instructions (Inst) executed during
TF-A boot stage before start of BL33 in RELEASE builds
for different versions is presented in the tables below,
where:
- C TF-A: existing TF-A C code;
- C musl: "lightweight code" C "implementation of the
standard library for Linux-based systems"
https://git.musl-libc.org/cgit/musl/tree/src/string/memset.c
- Asm Opt: assemler version from "Arm Optimized Routines"
project
https://github.com/ARM-software/optimized-routines/blob/
master/string/arm/memset.S
- Asm Linux: assembler version from Linux kernel
https://github.com/torvalds/linux/blob/master/arch/arm/lib/memset.S
- Asm TF-A: assembler version from this patch
Pramod Kumar [Wed, 5 Feb 2020 05:57:57 +0000 (11:27 +0530)]
lib: cpu: Check SCU presence in DSU before accessing DSU registers
The DSU contains system control registers in the SCU and L3 logic to
control the functionality of the cluster. If "DIRECT CONNECT" L3
memory system variant is used, there won't be any L3 cache,
snoop filter, and SCU logic present hence no system control register
will be present. Hence check SCU presence before accessing DSU register
for DSU_936184 errata.
arm_fpga: Add support to populate the CPU nodes in the DTB
At the moment BL31 dynamically discovers the CPU topology of an FPGA
system at runtime, but does not export it to the non-secure world.
Any BL33 user would typically looks at the devicetree to learn about
existing CPUs.
This patch exports a minimum /cpus node in a devicetree to satisfy
the binding. This means that no cpumaps or caches are described.
This could be added later if needed.
An existing /cpus node in the DT will make the code bail out with a
message.
Signed-off-by: Javier Almansa Sobrino <javier.almansasobrino@arm.com>
Change-Id: I589a2b3412411a3660134bdcef3a65e8200e1d7e
Anders Dellien [Sun, 23 Aug 2020 18:32:48 +0000 (19:32 +0100)]
plat/arm: Add dependencies to configuration files
This patch adds dependencies to the generated configuration
files that are included in the FIP. This fixes occasional
build errors that occur when the FIP happens to be built first.
Change-Id: I5a2bf724ba3aee13954403b141f2f19b4fd51d1b Signed-off-by: Anders Dellien <anders.dellien@arm.com>
This patch removes the assert that expects the SPMC PC
value to be same as BL32_BASE. This assumption is not
true for all platforms e.g. Tegra, and so will be removed
from the SPMD.
Platforms can always add this check to the platform files,
if required.
projects / arm-tf.git / log