feat(cpus): fix external LLC presence bit in Neoverse N3

Unlike Neoverse N2, Neoverse N3 incorporates the External LLC presence
bit in CPUECTLR2_EL1.SW_EXT_LLC. In addition, the default value is
ext

feat(cpus): fix external LLC presence bit in Neoverse N3

Unlike Neoverse N2, Neoverse N3 incorporates the External LLC presence
bit in CPUECTLR2_EL1.SW_EXT_LLC. In addition, the default value is
external LLC in Neoverse N3, so the bit will be cleared when
NEOVERSE_Nx_EXTERNAL_LLC is not enabled.

Change-Id: I1182aba5423e74748efd2571cc3817634ada748d
Signed-off-by: Younghyun Park <younghyunpark@google.com>

show more ...

fix(smccc): cleanup unused declaration

Remove check_wa_cve_2024_7881() declaration left behind by patch
fd04156eb792963cb21144063e421d074efa6386.

Signed-off-by: Arvind Ram Prakash <arvind.ramprakas

fix(smccc): cleanup unused declaration

Remove check_wa_cve_2024_7881() declaration left behind by patch
fd04156eb792963cb21144063e421d074efa6386.

Signed-off-by: Arvind Ram Prakash <arvind.ramprakash@arm.com>
Change-Id: Iae81e0bcbd2feb2b72a4e4e8fd27a27ae00c9cb4

show more ...

Merge changes Id38d6f1b,I5fcfe8dd,I7f3b50e5 into integration

* changes:
fix(cpus): inform the compiler that struct cpu_ops is aligned
refactor(el3-runtime): move the initialisation of the cpu_op

Merge changes Id38d6f1b,I5fcfe8dd,I7f3b50e5 into integration

* changes:
fix(cpus): inform the compiler that struct cpu_ops is aligned
refactor(el3-runtime): move the initialisation of the cpu_ops_ptr to C
fix(aarch32): make get_cpu_ops_ptr() PCS compliant

show more ...

fix(cpus): inform the compiler that struct cpu_ops is aligned

The only way to access a cpu_ops structure is through a pointer returned
from assembly so the compiler can't know its alignment and it m

fix(cpus): inform the compiler that struct cpu_ops is aligned

The only way to access a cpu_ops structure is through a pointer returned
from assembly so the compiler can't know its alignment and it must
assume the worst. As a result, it's scared to do 64 bit loads and must
do 8 single byte loads that it then can combine together.

Well, the cpu assembly macros take care to align the cpu_ops entries to
a word boundary so we can propagate that information to the structure
definition as well and removed the compiler's paranoia.

Change-Id: Id38d6f1b92527b8a414cfbb856a5a82c76a1b1a8
Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>

show more ...

feat(cpus): add support for Canyon CPU

Add basic CPU library code to support the Canyon CPU.

Change-Id: I82edc4384c4fe35ec2cf6b4bfd877a24ad8725dc
Signed-off-by: Icen Zeyada <Icen.Zeyada2@arm.com>
S

feat(cpus): add support for Canyon CPU

Add basic CPU library code to support the Canyon CPU.

Change-Id: I82edc4384c4fe35ec2cf6b4bfd877a24ad8725dc
Signed-off-by: Icen Zeyada <Icen.Zeyada2@arm.com>
Signed-off-by: Ryan Everett <ryan.everett@arm.com>
Signed-off-by: Min Yao Ng <minyao.ng@arm.com>
Signed-off-by: Aditya Deshpande <aditya.deshpande@arm.com>

show more ...

fix(cpus): workaround for Cortex-X4 erratum 3887999

Cortex-X4 erratum 3887999 is a Cat B erratum that applies
to all revisions <= r0p3 and is still open.

The erratum can be avoided by setting CPUAC

fix(cpus): workaround for Cortex-X4 erratum 3887999

Cortex-X4 erratum 3887999 is a Cat B erratum that applies
to all revisions <= r0p3 and is still open.

The erratum can be avoided by setting CPUACTLR2[22] to 1'b1 which will
disable linking multiple Non-Cacheable or Device GRE loads to the same
read request for the cache-line. This might have a significant
performance impact to Non-cacheable and Device GRE read bandwidth for
streaming scenarios

SDEN documentation:
https://developer.arm.com/documentation/SDEN-2432808/latest

Signed-off-by: Arvind Ram Prakash <arvind.ramprakash@arm.com>
Change-Id: I851746b7b430eac85184c8d402d1aa5bb3c94a8e

show more ...

fix(cpus): workaround for Cortex-X2 erratum 1934260

Cortex-X2 erratum 1934260 is a Cat B erratum that applies only
to revision r1p0 and is fixed in r2p0.

The workaround is to set CPUECTLR_EL1[25:18

fix(cpus): workaround for Cortex-X2 erratum 1934260

Cortex-X2 erratum 1934260 is a Cat B erratum that applies only
to revision r1p0 and is fixed in r2p0.

The workaround is to set CPUECTLR_EL1[25:18] to 0xFF. This
workaround will result in reduced performance for workloads
that benefit from write streaming.

SDEN documentation:
https://developer.arm.com/documentation/SDEN-1775100/latest

Change-Id: I180d38fee27175dc8ac5fa6726e5b71c3340285f
Signed-off-by: John Powell <john.powell@arm.com>

show more ...

fix(cpus): workaround for Cortex-X2 erratum 1901946

Cortex-X2 erratum 1901946 is a Cat B erratum that applies to
revision r1p0 and is fixed in r2p0.

The workaround is to set CPUACTLR4_EL1[15]. This

fix(cpus): workaround for Cortex-X2 erratum 1901946

Cortex-X2 erratum 1901946 is a Cat B erratum that applies to
revision r1p0 and is fixed in r2p0.

The workaround is to set CPUACTLR4_EL1[15]. This has a small
performance impact.

SDEN documentation:
https://developer.arm.com/documentation/SDEN-1775100/latest

Change-Id: I5a65db60f06982191994db49815419c4d72506cf
Signed-off-by: John Powell <john.powell@arm.com>

show more ...

feat(psci): check that CPUs handled a pabandon

Up to now PSCI assumed that if a pabandon happened then the CPU driver
will have handled it. This patch adds a simple protocol to make sure
that this i

feat(psci): check that CPUs handled a pabandon

Up to now PSCI assumed that if a pabandon happened then the CPU driver
will have handled it. This patch adds a simple protocol to make sure
that this is indeed the case. The chosen method is with a return value
that is highly unlikely on cores that are unaware of pabandon (x0 will
be primed with 1 and if used should be overwritten with the value of
CPUPWRCTLR_EL1 which should have its last bit set to power off and its
top bits RES0; the ACK value is chosen to be the exact opposite). An
alternative method would have been to add a field in cpu_ops, however
that would have required more major refactoring across many cpus and
would have taken up more memory on older platforms, so it was not
chosen.

Change-Id: I5826c0e4802e104d295c4ecbd80b5f676d2cd871
Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>

show more ...

feat(dsu): support power control and autonomous powerdown config

This patch allows platforms to enable certain DSU settings
to ensure memory retention and control over
cache power requests. We also

feat(dsu): support power control and autonomous powerdown config

This patch allows platforms to enable certain DSU settings
to ensure memory retention and control over
cache power requests. We also move the driver out of css
into drivers/arm. Platforms can configure the
CLUSTERPWRCTLR and CLUSTERPWRDN registers [1] to improve
power efficiency.

These registers enable finer-grained control of
DSU power state transitions, including
powerdown and retention.

IMP_CLUSTERPWRCTLR_EL1 provides:
- Functional retention: Allows configuration of the
duration of inactivity before the DSU uses
CLUSTERPACTIVE to request functional retention.

- Cache power request: These bits are output on
CLUSTERPACTIVE[19:16] to indicate to the power controller
which cache portions must remain powered.

IMP_CLUSTERPWRDN_EL1 includes:
- Powerdown: Triggers full cluster powerdown, including
control logic.

- Memory retention: Requests memory retention mode,
keeping L3 RAM contents while powering off
the rest of the DSU.

The DSU-120 TRM [2] provides the full field definitions,
which are used as references in the `dsu_driver_data` structure.

References:
[1]: https://developer.arm.com/documentation/100453/latest/
[2]: https://developer.arm.com/documentation/102547/0201/?lang=en

Signed-off-by: Arvind Ram Prakash <arvind.ramprakash@arm.com>
Change-Id: I2eba808b8f2a27797782a333c65dd092b03208fe

show more ...

Merge changes I1fae91a5,I54793492,I703f0e6e into integration

* changes:
fix(cpus): workaround for Cortex-A710 erratum 1917258
fix(cpus): workaround for Cortex-A710 erratum 1916945
fix(cpus): w

Merge changes I1fae91a5,I54793492,I703f0e6e into integration

* changes:
fix(cpus): workaround for Cortex-A710 erratum 1917258
fix(cpus): workaround for Cortex-A710 erratum 1916945
fix(cpus): workaround for Cortex-A710 erratum 1901946

show more ...

feat(errata): implement workaround for DSU-120 erratum 2900952

DSU Erratum 2900952 is a Cat B erratum that applies to some
DSU-120 implementations of revision r2p0 and is fixed in r2p1.
This erratum

feat(errata): implement workaround for DSU-120 erratum 2900952

DSU Erratum 2900952 is a Cat B erratum that applies to some
DSU-120 implementations of revision r2p0 and is fixed in r2p1.
This erratum is fixed in certain implementations of r2p0 which can be
determined by reading the IMP_CLUSTERREVIDR_EL1[1] register field
where a set bit indicates that the erratum is fixed in this part.

The workaround is to set the CLUSTERACTLR_EL1 bits [21:20] to 0x3
which ignores CBusy from the system interconnect and
setting CLUSTERACTLR_EL1 bit [8] to 1 to assert CBusy from DSU to
all the cores when DSU is busy.

SDEN: https://developer.arm.com/documentation/SDEN-2453103/1200/?lang=en

Signed-off-by: Arvind Ram Prakash <arvind.ramprakash@arm.com>
Change-Id: I87aa440ab5c35121aff703032f5cf7a62d0b0bb4

show more ...

fix(cpus): workaround for Cortex-A710 erratum 1901946

Cortex-A710 erratum 1901946 is a Cat B erratum that applies
to revision r1p0 and is fixed in r2p0.

The workaround is to set CPUACTLR4_EL1[15].

fix(cpus): workaround for Cortex-A710 erratum 1901946

Cortex-A710 erratum 1901946 is a Cat B erratum that applies
to revision r1p0 and is fixed in r2p0.

The workaround is to set CPUACTLR4_EL1[15]. This has a slight
performance impact.

SDEN documentation:
https://developer.arm.com/documentation/SDEN1775101

Change-Id: I703f0e6ee122e44a9bc284d90f1465039e3b40e4
Signed-off-by: John Powell <john.powell@arm.com>

show more ...

Merge "fix(errata): keep leading zeros in CVE ID prints" into integration

feat(cpus): update cpu_check_csv2 check

Update the cpu_check_csv2 logic to allow ID_AA64PFR0_EL1.CSV2
values up to 3. With the introduction of FEAT_CSV2_3,
the architectural limit for CSV2 has been

feat(cpus): update cpu_check_csv2 check

Update the cpu_check_csv2 logic to allow ID_AA64PFR0_EL1.CSV2
values up to 3. With the introduction of FEAT_CSV2_3,
the architectural limit for CSV2 has been extended,
making values from 0 to 3 valid.

Signed-off-by: Arvind Ram Prakash <arvind.ramprakash@arm.com>
Change-Id: I8473047ed4ad759b7b506161a76774ac21555d31

show more ...

fix(errata): keep leading zeros in CVE ID prints

The errata printing function would drop leading zeros on CVE
numbers so this updates the format string to make sure they are
printed. This is to conf

fix(errata): keep leading zeros in CVE ID prints

The errata printing function would drop leading zeros on CVE
numbers so this updates the format string to make sure they are
printed. This is to conform to the CVE naming convention where
ID numbers of less than 4 digits are prepended with 0s up to 4
digits.

This also updates a confusing comment indicating that leading
zeros could be used to work around a potential issue if CVE and
erratum IDs clash. Values with leading zeros will be interpreted
as octal numbers which is not desirable behavior so this should
not be recommended. Realistically, a CVE ID and erratum ID being
the same is *extremely* unlikely since CVE ID start over each
year and are 4-5 digits for Arm, and Errata IDs are 6-7 digits.

Change-Id: Idf2be50cea6828a3d30c6e58fda477ec1398bc7c
Signed-off-by: John Powell <john.powell@arm.com>

show more ...

Merge changes I005586ef,I0d4d74bc into integration

* changes:
fix(cpufeat): replace "bti" mnemonic with hint instructions
fix(cpufeat): improve xpaci wrapper

fix(cpufeat): replace "bti" mnemonic with hint instructions

Older GNU binutils version require to specify at least "armv8.5-a" for
the ARM architecture revision to accept "bti" instructions in the
a

fix(cpufeat): replace "bti" mnemonic with hint instructions

Older GNU binutils version require to specify at least "armv8.5-a" for
the ARM architecture revision to accept "bti" instructions in the
assembly code. Binutils v2.35 have relaxed this, since "bti" is in the
hint space, so is ignored on older cores and does NOT require a BTI
enabled core to execute.

To not exclude those older binutils versions (as shipped with Ubuntu
20.04), use the "hint" encoding for the "bti" instructions, which are
accepted regardless of the minimum architecture revision. Hide this
encoding in a macro, to make the "bti" usage more readable in the
source code.

Change-Id: I005586efd8974a3f2c7202896c881bb5fed07eea
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

show more ...

refactor(cpus): optimize CVE checking

This patch replaces the use of EXTRA functions
with using erratum entries check
to verify CVE mitigation application for some of
the SMCCC_ARCH_WORKAROUND_* cal

refactor(cpus): optimize CVE checking

This patch replaces the use of EXTRA functions
with using erratum entries check
to verify CVE mitigation application for some of
the SMCCC_ARCH_WORKAROUND_* calls.

Previously, EXTRA functions were individually implemented for
each SMCCC_ARCH_WORKAROUND_*, an approach that becomes unmanageable
with the increasing number of workarounds.
By looking up erratum entries for CVE check, the process is streamlined,
reducing overhead associated with creating and
maintaining EXTRA functions for each new workaround.

New Errata entries are created for SMC workarounds and
that is used to target cpus that are uniquely impacted
by SMC workarounds.

Signed-off-by: Arvind Ram Prakash <arvind.ramprakash@arm.com>
Change-Id: I873534e367a35c99461d0a616ff7bf856a0000af

show more ...

refactor(cpus): move errata check to common code

This patch centralizes some of the Errata ABI code
that could be used for checking if an Errata has been applied
to cpu library since the function is

refactor(cpus): move errata check to common code

This patch centralizes some of the Errata ABI code
that could be used for checking if an Errata has been applied
to cpu library since the function is mostly generic.

Signed-off-by: Arvind Ram Prakash <arvind.ramprakash@arm.com>
Change-Id: I2c6d4468f7125d4d99ccdebc5ea8f9e4390360cc

show more ...

feat(psci): remove cpu context init by index

Currently, the calling core (meaning the core which received the call to
CPU_ON or the powerdown path of CPU_SUSPEND on the same core) is in
charge of in

feat(psci): remove cpu context init by index

Currently, the calling core (meaning the core which received the call to
CPU_ON or the powerdown path of CPU_SUSPEND on the same core) is in
charge of initialising the context for the waking core (the warmboot
entrypoint for both). This is convenient because the calling core can
write the context while in coherency and the waking core will only need
the context after its entered coherency. This avoids any cache
maintenance and makes communication simple.

However, this has 3 main problems:
a) asymmetric feature support is problematic - the calling core has no
way of knowing the feature set of the waking core. If the two
diverge, the architectural feature discovery via ID registers breaks
down. We've thus far "fixed" this on a case by case basis which
doesn't scale and introduces redundancy.

b) powerdown abandon (pabandon) introduces a contradiction - the calling
core has to initialise the context for when the core wakes up, but
should the core not powerdown it needs its old context intact. The only
way to work around this is by keeping two copies of context which
incurs a runtime and memory overhead.

c) cm_prepare_el3_exit[_ns]() doesn't have access to the entrypoint but needs
it to make initialisation decisions. We can infer some of this from
registers that have already been written but this is awkwardly
limiting for what we can do. This also necessitates the split from
the context initialisation.

We can solve all three by a making a core be in full ownership of its
own context. The calling core then only writes entrypoint information
and nothing else. The waking core then initialises its own context as it
sees fit with full knowledge of the whole picture.

The only tricky bit is cache coherency - the waking core has to be able
to coherently observe its new entrypoint. Calling cores will write to
the shared region with coherent caches on. If we make sure to read the
context only after the waking core has entered coherency, then we can
avoid cache operations and let hardware handle everything.

We can skip the spsr check for FEAT_TCR2 as it doesn't make a
difference. We can also skip enabling it twice from generic code.

Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>
Signed-off-by: Manish Pandey <manish.pandey2@arm.com>
Change-Id: I86e7fe8b698191fc3b469e5ced1fd010f8754b0e

show more ...

fix(cpus): fix clang compilation issue

A potential problem with clang version < 17 can cause resolving nested
'cfi_startproc' to fail compilation.

So add a variant of check_errara/reset_macros that

fix(cpus): fix clang compilation issue

A potential problem with clang version < 17 can cause resolving nested
'cfi_startproc' to fail compilation.

So add a variant of check_errara/reset_macros that is compatible with
clang version < 17 to ignore `cfi_startproc` and `cfi_endproc`.

This wouldn't cause any performance issue and will not affect any
functional behaviour.

Change-Id: I46147af2dd0accd5be14ddb26dea03bb2f87cba8
Signed-off-by: Govindraj Raja <govindraj.raja@arm.com>

show more ...

fix(cpus): remove errata setting PF_MODE to conservative

The erratum titled “Disabling of data prefetcher with outstanding
prefetch TLB miss might cause a deadlock” should not be handled within
TF-A

fix(cpus): remove errata setting PF_MODE to conservative

The erratum titled “Disabling of data prefetcher with outstanding
prefetch TLB miss might cause a deadlock” should not be handled within
TF-A. The current workaround attempts to follow option 2 but
misapplies it. Specifically, it statically sets PF_MODE to
conservative, which is not the recommended approach. According to the
erratum documentation, PF_MODE should be configured in conservative
mode only when we disable data prefetcher however this is not done
in TF-A and thus the workaround is not needed in TF-A.

The static setting of PF_MODE in TF-A does not correctly address the
erratum and may introduce unnecessary performance degradation on
platforms that adopt it without fully understanding its implications.

To prevent incorrect or unintended use, the current implementation of
this erratum workaround should be removed from TF-A and not adopted by
platforms.

List of Impacted CPU's with Errata Numbers and reference to SDEN -

Cortex-A78 - 2132060 - https://developer.arm.com/documentation/SDEN1401784/latest
Cortex-A78C - 2132064 - https://developer.arm.com/documentation/SDEN-2004089/latest
Cortex-A710 - 2058056 - https://developer.arm.com/documentation/SDEN-1775101/latest
Cortex-X2 - 2058056 - https://developer.arm.com/documentation/SDEN-1775100/latest
Cortex-X3 - 2070301 - https://developer.arm.com/documentation/SDEN2055130/latest
Neoverse-N2 - 2138953 - https://developer.arm.com/documentation/SDEN-1982442/latest
Neoverse-V1 - 2108267 - https://developer.arm.com/documentation/SDEN-1401781/latest
Neoverse-V2 - 2331132 - https://developer.arm.com/documentation/SDEN-2332927/latest

Change-Id: Icf4048508ae070b2df073cc46c63be058b2779df
Signed-off-by: Govindraj Raja <govindraj.raja@arm.com>

show more ...

fix(errata): workaround for Cortex-A510 erratum 2971420

Cortex-A510 erratum 2971420 applies to revisions r0p1, r0p2, r0p3,
r1p0, r1p1, r1p2 and r1p3, and is still open.

Under some conditions, data

fix(errata): workaround for Cortex-A510 erratum 2971420

Cortex-A510 erratum 2971420 applies to revisions r0p1, r0p2, r0p3,
r1p0, r1p1, r1p2 and r1p3, and is still open.

Under some conditions, data might be corrupted if Trace Buffer
Extension (TRBE) is enabled. The workaround is to disable trace
collection via TRBE by programming MDCR_EL3.NSTB[1] to the opposite
value of SCR_EL3.NS on a security state switch. Since we only enable
TRBE for non-secure world, the workaround is to disable TRBE by
setting the NSTB field to 00 so accesses are trapped to EL3 and
secure state owns the buffer.

SDEN: https://developer.arm.com/documentation/SDEN-1873361/latest/

Signed-off-by: John Powell <john.powell@arm.com>
Change-Id: Ia77051f6b64c726a8c50596c78f220d323ab7d97

show more ...

fix(cpus): workaround for Cortex-A715 erratum 2804830

Cortex-A715 erratum 2804830 applies to r0p0, r1p0, r1p1 and r1p2,
and is fixed in r1p3.

Under some conditions, writes of a 64B-aligned, 64B gra

fix(cpus): workaround for Cortex-A715 erratum 2804830

Cortex-A715 erratum 2804830 applies to r0p0, r1p0, r1p1 and r1p2,
and is fixed in r1p3.

Under some conditions, writes of a 64B-aligned, 64B granule of
memory might cause data corruption without this workaround. See SDEN
for details.

Since this workaround disables write streaming, it is expected to
have a significant performance impact for code that is heavily
reliant on write streaming, such as memcpy or memset.

SDEN: https://developer.arm.com/documentation/SDEN-2148827/latest/

Change-Id: Ia12f6c7de7c92f6ea4aec3057b228b828d48724c
Signed-off-by: John Powell <john.powell@arm.com>

show more ...