* changes:
fix(qemu): statically allocate bitlocks array
feat(qemu): update for renamed struct memory_bank
feat(fvp): increase GPT PPS to 1TB
feat(gpt): statically allocate bitlocks array
chore(gpt): define PPS in platform header files
feat(fvp): allocate L0 GPT at the top of SRAM
feat(fvp): change size of PCIe memory region 2
feat(rmm): add PCIe IO info to Boot manifest
feat(fvp): define single Root region
gpt_runtime_init() now takes the bitlock array's address and size as
argument. Rather than reserving space at the end of the L0 GPT for
storing bitlocks, allocate a static array and pass its address to
gpt_runtime_init(). This frees up a little bit of space formerly
reserved for alignment of the GPT.
Change-Id: I48a1a2bc230f64e13e3ed08b18ebdc2d387d77d0
Signed-off-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
To allow for generic handling of a wakeup, this hook is no longer
expected to call wfi itself. Update the name everywhere to reflect this
expectation so that future platform implementers don't get misled.
Change-Id: Ic33f0b6da74592ad6778fd802c2f0b85223af614
Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>
This change introduces the hob support for both qemu platforms (virt and
sbsa).
As the hob list feature relies on transfer list, the transfer list
support is promoted to common qemu build configuration. The platforms
specific definitions are updated accordingly.
Change-Id: I473d83388fe95408d34515bf7bcbdd64ce4e777d
Signed-off-by: Kun Qin <kuqin@microsoft.com>
Add the necessary foundation to support Arm's RME extension on the SBSA
reference architecture.
Change-Id: If5a63ed0015cb33fcae367ff2cded811bbdc1e54
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
When RME is enabled on SBSA, the RMM is located at the start of the NS
RAM and the device tree after it. This patch adjust the DT memory start
address so that anyone reading it has an accurate view of the system
configuration.
Change-Id: I32ca63a78d68831faf2c65ad60a45c841b7cbada
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
When RME is enabled, (1) the RMM is installed at the base of system RAM,
(2) the base of the system RAM is shifted upward, after the RMM and (3)
the device tree is relocated to the new system RAM base.
This patch relocates the device tree to the new system RAM base before
the RMM is installed in RAM. From there, other accesses to the device
tree are using the new location.
Change-Id: I0cb4e060ca33a11becd78fe48fab4dc76f0b484b
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
When RME is enabled the RMM is placed at the bottom of the NS RAM,
meaning that NS_DRAM0_BASE has to be located after that.
This patch disscociates the base of the NS RAM as defined by QEMU by
introducing a new define, PLAT_QEMU_DRAM0_BASE. An offset can be added
to that new define when the software's view of the base memory need to
differ from QEMU.
No change in functionality.
Change-Id: I887f9993d5a61896352cfff17e0d92e2c2b9030a
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Following what was done for:
f465ac2210 ("fix(qemu): increase max FIP size")
increase the size of the FIP image to take up the remaining
space in FLASH0. That way the RMM image can also be added
to the FIP.
Change-Id: I89bba36f751468e99241f1c20b51c48fe06d8229
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Move all DT related functions to file sbsa_platform_dt.c so that clients
other than SIP SVC can use the funtionality. At the same time, make all
functions that don't need outside visibility static.
No change in functionality.
Change-Id: I9bce730c8f9e2b827937466f4432ecfa74c35675
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Creating accessor functions to access information held by struct
qemu_platform_info. That way the code that is relevant to fetching
information from the device tree can be taken out of sbsa_sip_svc.c and
placed in a file where other client can use the information it provides.
No change in functionality.
Change-Id: I989952ee6d15e1436549002dd7c7767c745ea297
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
There is no relation between the name of function sip_svc_init() and
what it does. As such rename it to something more appropriate and move
it to a header that make sense.
No change in functionality.
Change-Id: I7bd78b1fe70e2930c395ef0a097bfad3b1e55d3a
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Move structure declaration related to the DT to their own header. That
way they can be reused by other files. At the same time, typedefs are
removed and structure names prepended with "platform_" to avoid clashing
with other structure declarations available in the system.
No change in functionality.
Change-Id: If67a141cc7441b0636af774d7edfe51cf8034a11
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Rename struct dynamic_platform_info to qemu_platform_info and properly
declare a variable name "dynamic_platform_info". That way structures
related to the device tree can be moved out of sbsa_sip_svc.c.
No change in functionality.
Change-Id: I1af39047af96ae02f3b8eecda6cb67508f14d37a
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
When building SBSA, using DT functions from fdt_wrappers.c produces a
linker error. Adding:
BL2_SOURCES += ${FDT_WRAPPERS_SOURCES}
fixes the problem. Since the same inclusion would be present in both
qemu/platform.mk and qemu_sbsa/platform.mk, do the changes in
qemu/common/common.mk.
Change-Id: I775b06c1741f6618813c5e1d2c64cdc1888d8519
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
We have fdt_read_uint32_default() function which allows us to use less
temporary variables. Let make use of it where applicable.
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Change-Id: I6fc8a87d5aac427703fd3c8b689e153ed58fa8b7
We add the support for adding cpus/topology to device tree in sbsaQemu
platform, and we can get this information via SMC calls:
- counting the number of sockets
- counting the number of clusters in one socket
- counting the number of cores in one cluster
- counting the number of threads in one core
Signed-off-by: Xiong Yining <xiongyining1480@phytium.com.cn>
Change-Id: I0059a5c7bb7055aba1aa5ec5bfd0ec78801874f8
From QEMU change:
> In previous versions of the Arm architecture, the frequency of the
> generic timers as reported in CNTFRQ_EL0 could be any IMPDEF value,
> and for QEMU we picked 62.5MHz, giving a timer tick period of 16ns.
> In Armv8.6, the architecture standardized this frequency to 1GHz.
This change stops TF-A from hardcoding 62.5MHz frequency. Instead value
stored in CNTFRQ_EL0 would be used. As a result we get 62.5MHz on older
cores and 1GHz on newer ones.
Change-Id: I7d414ce6d3708e598bbb5a6f79eb2d4ec8e15ac4
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
As a part of removing DeviceTree from EDK2, we move functions to TF-A:
- counting the number of memory nodes
- checking NUMA node id
- checking the memory address
Signed-off-by: Xiong Yining <xiongyining1480@phytium.com.cn>
Signed-off-by: Chen Baozi <chenbaozi@phytium.com.cn>
Change-Id: Ib7bce3a65c817a5b3bef6c9e0a459c7ce76c7e35
Coverity Scan reminded that we need to take care of MPIDR properly.
We need to make sure that we get MPIDR values from QEMU.
No MPIDR == panic() in case which should not happen.
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Change-Id: Idb5fe7d958f0bcecd3d66a643743f478538f4a8b
QEMU provides us with minimal information about hardware platform using
minimalistic DeviceTree. This is not a Linux DeviceTree. It is not even
a firmware DeviceTree.
Change-Id: I7b6cc5f53a4f78a9ed69bc7fc2fa1a69ea65428d
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
We want to remove use of DeviceTree from EDK2. So we move
functions to TF-A:
- counting cpu cores
- checking NUMA node id
- checking MPIDR
And then it gets passed to EDK2 via SMC calls.
Change-Id: I1c7fc234ba90ba32433b6e4aa2cf127f26da00fd
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
GICV2_G0_FOR_EL3 variable is only for GICv2 platforms.
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Change-Id: Ibf9376caabbc05ceef4f870d816e6c60a344f895
RME patch series shown that we can build larger BL1 than we can run:
NOTICE: Booting Trusted Firmware
NOTICE: BL1: v2.9(debug):v2.9.0-736-g08548888a
NOTICE: BL1: Built : 12:10:39, Sep 18 2023
INFO: BL1: RAM 0x3ffee000 - 0x3fffb000
INFO: BL1: Loading BL2
WARNING: Firmware Image Package header check failed.
RME pushed debug build BL1 over 0x8000 in size.
This exposed an error where FIP_BASE (supposed to be at BL1_SIZE offset
from start of flash) was actually 0x8000 and not 0x12000.
Make sure we have space for BL1 by deriving FIP_BASE from it.
Note: this is a breaking change for edk2 FD image generation, which had
similarly hardcoded a 0x8000 offset. These images must be updated in
lock-step.
Change-Id: I8a1a85e82319945a4412c424467d818d5b6e4ecd
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
MULTI_CONSOLE_API have been removed long time ago by commit 5b6ebeec9c
("Remove MULTI_CONSOLE_API flag and references to it") that's why remove
references in platform.mk files and also in one rst which is not valid
anymore.
Change-Id: I45f8e7db0a14ce63de62509100d8159b7aca2657
Signed-off-by: Michal Simek <michal.simek@amd.com>
There are some variables to enable/disable options. Let keep them at top
of file.
Change-Id: I108dd814557b6c713aba0d73a52148c766079c8b
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
SPM_MM is not compatible with ENABLE_SVE_FOR_NS and breaks build early:
> Including SPM Management Mode (MM) makefile
> services/std_svc/spm/spm_mm/spm_mm.mk:14: *** "Error: SPM_MM is not compatible with ENABLE_SVE_FOR_NS". Stop.
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Change-Id: Iabe181647fce00a432ae11dc4599b71619364c24
Handle coherency in one place for AArch64 mode.
Change-Id: Id3678a8f478e5ef731c81c0df30059000e380758
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Enable SVE, SME, RNG, FGT in one place.
qemu gains FGT (needed for 'max' cpu to boot Linux)
qemu_sbsa gains RNG
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Change-Id: I2e8f971ef3e42d9ebe9f20641b288cc8c40f806a
Move BL31 source list into common file.
Change-Id: Iaa27cfd8f87b691728379c87a6ff6331e87951e1
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Move BL1 and BL2 source list into common file.
Change-Id: I8f9a835f6cd1c5d67728a071860173f80f03c84e
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Keep list of supported cpu cores in one place for both platforms.
qemu_sbsa does not handle some of them but with 256MB firmware space it
does not matter.
Change-Id: I5b8f7d18dc903e86e0cc7babbc2fb3f26a1bfdfa
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Move libfdt includes into common file and use definitions from them.
Change-Id: Ic4fe784fdbedcf5e9e3804a633fcac68464f38a6
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Add support to qemu "neoverse-v1" cpu for "qemu_sbsa" ('sbsa-ref')
platform.
Change-Id: Id710e2b960e7938d2dbe7a88d9e158a7009fc3d1
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Read data from DeviceTree provided by QEMU, provide via SMC
to the next firmware level.
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Change-Id: I95c5f00ab2cca3b5fda122dcc8d7704a7a82059b
QEMU provides GIC information in DeviceTree (on platform version 0.1+).
Read it and provide to next firmware level via SMC.
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Change-Id: I383919bd172acc8873292a0c5e4469651dc96fb9
QEMU provides platform version information via DT. We want to use it
in firmware to handle differences between platform versions.
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Change-Id: I8def66dac9dd5d7ab0e459baa40e27a11b65f0ba
QEMU 7.2+ has FEAT_FGT support added to 'max' cpu.
So let's enable it to make Debian 'bookworm' kernel boot on sbsa-ref/max setup.
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Change-Id: I49fb3e742b69ce7be5666e0144525dde21a68238
Currently, TF-A supports three states for feature flags:
0: FEAT_DISABLED
1: FEAT_STATE_ALWAYS ( for fixed/real platforms)
2: FEAT_STATE_CHECK ( for configurable platforms)
to meet the feature detection requirements dynamically, mainly
targetting configurable/Fixed Virtual platforms.
With this mechanism in place, we are refactoring all the existing
feature flags to the FEAT_STATE_CHECK option(=2), including
FEAT_SVE explicitly for FVPs.
SVE Patch Reference:
https://review.trustedfirmware.org/c/TF-A/trusted-firmware-a/+/19888/25
This newly introduced change, breaks the existing behaviour especially
for virtual platforms, who have set the ENABLE_SVE_FOR_NS flag to 1.
Moving ahead, we advise the platforms to take the following steps while
enabling the features:
1. If the platform is configurable (virtual), and want to ensure feature
detection happens dynamically at runtime, set the build flags to
FEAT_STATE_CHECK(=2).
2. For real(fixed) platforms, depending on the features supported by the
hardware and platform wants to enable it, platforms could set build
flags to FEAT_STATE_ALWAYS(=1).
(Note: Only the non-secure world enablement related build flags have
been refactored to take the values within 0 to 2. As earlier Secure
world enablement flags will still remain boolean.)
Henceforth, in order to keep it aligned with this tri-state mechanism,
changing the qemu platform default to the now supported dynamic
option(=2), so the right decision can be made by the code at runtime.
Change-Id: Icc95b8b872378b7874d4345b631adfc314e4dada
Signed-off-by: Jayanth Dodderi Chidanand <jayanthdodderi.chidanand@arm.com>
Add support to qemu "neoverse-n1" cpu for "qemu_sbsa" ('sbsa-ref')
platform.
Signed-off-by: Chen Baozi <chenbaozi@phytium.com.cn>
Change-Id: I4620e879c71115451ae91a1643812d89ec7c071f
Some of our specialized sections are not prefixed with the conventional
period. The compiler uses input section names to derive certain other
section names (e.g. `.rela.text`, `.relacpu_ops`), and these can be
difficult to select in linker scripts when there is a lack of a
delimiter.
This change introduces the period prefix to all specialized section
names.
BREAKING-CHANGE: All input and output linker section names have been
prefixed with the period character, e.g. `cpu_ops` -> `.cpu_ops`.
Change-Id: I51c13c5266d5975fbd944ef4961328e72f82fc1c
Signed-off-by: Chris Kay <chris.kay@arm.com>
Commit 337ff4f1dd enabled SVE/SME for
qemu platform. Let do the same for qemu-sbsa one too.
With this change I can boot Debian 'bookworm' installed using Max cpu.
Info from referenced commit:
Starting with QEMU v3.1.0 (Dec 2018), QEMU's TCG emulation engine supports
the SVE architecture extension. In QEMU v7.1.0 (Aug 2022) it also gained
SME support.
As it stands today, running TF-A under QEMU with "-cpu max" makes Linux
hang, because SME and SVE accesses trap to EL3, but are never handled
there. This is because the Linux kernel sees the SVE or SME feature bits,
and assumes firmware has enabled the feature for lower exception levels.
This requirement is described in the Linux kernel booting protocol.
Enable those features in the TF-A build, so that BL31 does the proper
EL3 setup to make the feature usable in non-secure world.
We check the actual feature bits before accessing SVE or SME registers,
so this is safe even for older QEMU version or when not running with
-cpu max. As SVE and SME are AArch64 features only, do not enable them
when building for AArch32.
Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Change-Id: I9ea1f91e6b801218d944e8a7d798d5ae568ed59a
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
Add support to qemu "max" cpu for both "qemu" ('virt') and
"qemu_sbsa" ('sbsa-ref') platforms.
Change-Id: I36e45c0a3c4e30ba546d2a3cb44dfef11a680305
Signed-off-by: Leif Lindholm <leif@nuviainc.com>
This commit fixes the wrong memory type, secure NOR flash
shall be mapped as MT_DEVICE.
Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org>
Change-Id: I9c9ed51675d84ded675bb56b2e4ec7a08184c602
sbsa-ref in QEMU may create up to 512 cores.
This commit prepares the MP information to support 512 cores.
The number of xlat tables for spm_mm is also increased.
Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org>
Change-Id: I2788eaf6d14e188e9b5d1102d359b2899e02df7c
This allows PSCI in TF-A to signal platform power states to QEMU
via a controller in secure space.
This required a sbsa-ref specific version of PSCI functions for the
platform. Also adjusted the MMU range to also include the new EC.
Add a new MMU region for the embedded controller and increase the
size of xlat tables by one for the new region.
Signed-off-by: Graeme Gregory <graeme@nuviainc.com>
Change-Id: Iece8a88947f11e82ab8988e460a8a66ad175a5ee
sbsa-ref in QEMU creates clusers of 8 cores, it may create up to 512
cores in upto 64 clusters. Implement a qemu_sbsa specific topology file
and increase the BL31_SIZE to accommodate the bigger table sizes. Change
platform_def.h for new topology. Correct PLATFORM_CPU_PER_CLUSTER_SHIFT so
plat_helpers.S calculates correct result.
Signed-off-by: Graeme Gregory <graeme@nuviainc.com>
Change-Id: Idc5d70394c0956b759ad2c86f9fda8f293f2cfa7
