Neoverse V2 erratum 2719105 is a Cat B erratum that applies to all
revisions <= r0p1 and is fixed in r0p2.
The erratum is avoided by setting CPUACTLR2_EL1[0] to 1 to force
PLDW/PFRM ST to behave like PLD/PRFM LD and not cause invalidations
to other PE caches. There might be a small performance degradation
to this workaround for certain workloads that share data.
SDEN documentation:
https://developer.arm.com/documentation/SDEN2332927/latest
Signed-off-by: Bipin Ravi <bipin.ravi@arm.com>
Change-Id: Id026edcb7ee1ca93371ce0001d18f5a8282c49ba
Neoverse V2 erratum 2331132 is a Cat B erratum that applies to all
revisions <= r0p2 and is still open. The workaround is to write the
value 4'b1001 to the PF_MODE bits in the IMP_CPUECTLR2_EL1 register
which will place the data prefetcher in the most conservative mode
instead of disabling it.
SDEN documentation:
https://developer.arm.com/documentation/SDEN2332927/latest
Signed-off-by: Bipin Ravi <bipin.ravi@arm.com>
Change-Id: Ic6c76375df465a4ad2e20dd7add7037477d973c1
Cortex-X3 erratum 2742421 is a Cat B erratum that applies to
all revisions <= r1p1 and is fixed in r1p2. The workaround is to
set CPUACTLR5_EL1[56:55] to 2'b01.
SDEN documentation:
https://developer.arm.com/documentation/2055130/latest
Signed-off-by: Sona Mathew <sonarebecca.mathew@arm.com>
Change-Id: Idadd323e419739fe909b9b68ea2dbe857846666b
Neoverse N2 erratum 2009478 is a Cat B erratum that applies to
revision r0p0 and is fixed in r0p1. The workaround is to clear
the ED bit for all core error records before setting the PWRDN_EN
bit in CPUPWRCTLR_EL1 to request a power down.
SDEN documentation:
https://developer.arm.com/documentation/SDEN1982442/latest
Signed-off-by: Bipin Ravi <bipin.ravi@arm.com>
Change-Id: Ic5ef58c9e795b90026af1d2b09edc0eea3ceee51
Adding basic CPU library code to support Nevis CPU
Change-Id: I399cc9b7b2d907b02b76ea2a3e5abb54e28fbf6c
Signed-off-by: Juan Pablo Conde <juanpablo.conde@arm.com>
* changes:
refactor(cpus): convert the Cortex-A57 to use cpu helpers
refactor(cpus): convert the Cortex-A57 to use the errata framework
refactor(cpus): reorder Cortex-A57 errata by ascending order
refactor(cpus): add Cortex-A57 errata framework information
refactor(cpus): convert the Cortex-A53 to use cpu helpers
refactor(cpus): convert the Cortex-A53 to use the errata framework
refactor(cpus): reorder Cortex-A53 errata by ascending order
This involves replacing:
* the reset_func with the standard cpu_reset_func_{start,end} to apply
errata automatically
* the <cpu>_errata_report with the errata_report_shim to report errata
automatically
...and for each erratum:
* the prologue with the workaround_<type>_start to do the checks and
framework registration automatically
* the epilogue with the workaround_<type>_end
* the checker function with the check_erratum_<type> to make it more
descriptive
It is important to note that the errata workaround sequences remain
unchanged and preserve their git blame.
At this point the binary output of all errata was checked with the
script from commit 19136. All reported discrepancies involve errata
with no workaround in the cpu file or errata that did not previously
have a workaround function and now do. The non temporal hint erratum has
been converted to a numeric erratum.
Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>
Change-Id: Ib321ab01362c5954fe78e7349229c1437b3da847
Errata report order is enforced to be in ascending order. To achieve
this with the errata framework this has to be done at the definition
level.
Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>
Change-Id: Ia98976797fc0811f30c7dbf714e94b36e3c2263e
Also, convert checker functions of errata which are enabled for all cpu
revisions to report correctly in preparation of the errata ABI.
Although the script from commit 250919 was used to check that errata
code did not change, this CPU only loosely adhered to convention and its
output was not particularly useful. Nevertheless, the discrepancies were
manually verified. All errata have been checked that they get invoked.
Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>
Change-Id: I988db6e7b6d1732f1d2258dbdf945cb475781894
This involves replacing:
* the reset_func with the standard cpu_reset_func_{start,end} to apply
errata automatically
* the <cpu>_errata_report with the errata_report_shim to report errata
automatically
...and for each erratum:
* the prologue with the workaround_<type>_start to do the checks and
framework registration automatically
* the epilogue with the workaround_<type>_end
* the checker function with the check_erratum_<type> to make it more
descriptive
It is important to note that the errata workaround and checking
sequences remain unchanged and preserve their git blame.
Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>
Change-Id: I30556f438859d17f054cb6bc96f3069b40474b58
Errata report order is enforced to be in ascending order. To achieve
this with the errata framework this has to be done at the definition
level.
Also rename the disable_non_temporal_hint to its erratum number to
conform to convention.
Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>
Change-Id: Id474872afebf361ab3d21c454ab3624db8354045
The original powerdown function for Gelas included SME disabling
instructions but did not check for the presence of SME before disabling.
This could lead to unexpected beaviors. This patch adds that check so
the feature is disabled only if it is present.
Change-Id: I582db53a6669317620e4f72a3eac87525897d3d0
Signed-off-by: Juan Pablo Conde <juanpablo.conde@arm.com>
MTE and common system registers are the last remaining EL2 save/restores
in assembly. Convert them to C, like all the others.
Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>
Change-Id: If690f792e70b97fd4b4cd5f43847a71719b128f1
This patch adds the necessary CPU library code to support the Gelas CPU
Change-Id: I13ec4a8bb7055c1ebd0796a4a1378983d930fcb3
Signed-off-by: Juan Pablo Conde <juanpablo.conde@arm.com>
This involves replacing:
* the reset_func with the standard cpu_reset_func_{start,end} to apply
errata automatically
* the <cpu>_errata_report with the errata_report_shim to report errata
automatically
...and for each erratum:
* the prologue with the workaround_<type>_start to do the checks and
framework registration automatically
* the epilogue with the workaround_<type>_end
* the checker function with the check_erratum_<type> to make it more
descriptive
It is important to note that the errata workaround sequences remain
unchanged and preserve their git blame.
Testing was conducted by:
* Manual comparison of disassembly of converted functions with non-
converted functions
aarch64-none-elf-objdump -D <trusted-firmware-a with conversion>/build/fvp/release/bl31/bl31.elf
vs
aarch64-none-elf-objdump -D <trusted-firmware-a clean repo>/build/fvp/release/bl31/bl31.elf
* Build for release with all errata flags enabled and run default tftf
tests
CROSS_COMPILE=aarch64-none-elf- make PLAT=fvp DEBUG=0 \
CTX_INCLUDE_AARCH32_REGS=0 HW_ASSISTED_COHERENCY=1 USE_COHERENT_MEM=0 \
BL33=./../tf-a-tests/build/fvp/release/tftf.bin \
WORKAROUND_CVE_2022_23960=1 ERRATA_ABI_SUPPORT=1 all fip
* Build for debug with all errata enabled and step through ArmDS
at reset to ensure all functions are entered.
Change-Id: I34e27e468d4f971423a03a95a4a52f4af8bd783a
Signed-off-by: Arvind Ram Prakash <arvind.ramprakash@arm.com>
Convert Neoverse V2 to use CPU helpers, in this case that's
only two spots.
Change-Id: Icd250f92974e8a50c459038de7644a2e68007589
Signed-off-by: Moritz Fischer <moritzf@google.com>
Signed-off-by: Arvind Ram Prakash <arvind.ramprakash@arm.com>
For V2, this involves replacing:
- The reset_func with the standard cpu_reset_func_{start,end}
to apply errata automatically
- The <cpu>_errata_report with the errata_report_shim to
report errata automatically
And for each erratum:
- The prologue with the workaround_<type>_start to do the checks and
framework registration automatically at reset or runtime
- The epilogue with the workaround_<type>_end
- The checker function with the check_erratum_<type> to check whether
the erratum applies on the revision of the CPU.
It is important to note that the errata workaround sequences remain
unchanged and preserve their git blame.
Testing was conducted by:
* Manual comparison of disassembly of converted functions with non-
converted functions
aarch64-none-elf-objdump -D <trusted-firmware-a with conversion>/build/fvp/release/bl31/bl31.elf
vs
aarch64-none-elf-objdump -D <trusted-firmware-a clean repo>/build/fvp/release/bl31/bl31.elf
* Build for release with all errata flags enabled and run default tftf
tests
CROSS_COMPILE=aarch64-none-elf- make PLAT=fvp CTX_INCLUDE_AARCH32_REGS=0 \
HW_ASSISTED_COHERENCY=1 USE_COHERENT_MEM=0 \
BL33=./../tf-a-tests/build/fvp/debug/tftf.bin \
ERRATA_V2_2801372 WORKAROUND_CVE_2022_23960=1 ERRATA_ABI_SUPPORT=1 all fip
* Build for debug with all errata enabled and step through ArmDS
at reset to ensure all functions are entered.
Change-Id: Ic968844d6aabea3867189d747769ced8faa87e56
Signed-off-by: Moritz Fischer <moritzf@google.com>
Signed-off-by: Arvind Ram Prakash <arvind.ramprakash@arm.com>
Not including the proper CPU file can halt execution at the
reset_handler since the cpu_ops obtained will be invalid and therefore
the cpu reset function will be invalid too, unless SUPPORT_UNKNOWN_MPID
is enabled.
This patch adds an assert to check for the validity of the obtained
cpu_ops object and will display an error if the object is invalid.
Change-Id: I0e1661745e4a692aab5e910e110c2de0caf64f46
Signed-off-by: Thaddeus Serna <Thaddeus.Gonzalez-Serna@arm.com>
Similar to Cortex-A53, the AArch32-only Cortex-A7/12/15/17 have an
(optional) integrated L2 cache that might need to be flushed if the
whole cluster is powered down. However, unlike Cortex-A53 there is
currently no L2 cache flush in the cluster_pwr_dwn implementation for
some reason. This causes problems if there is unwritten data left in
the L2 cache during a cluster power off.
Fix this by adding the L2 cache flush similar to cortex_a53.S.
Change-Id: Icd087bef9acff11e03edcaa0d26dd8b8e30796b7
Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
Align entire TF-A to use Arm in copyright header.
Change-Id: Ief9992169efdab61d0da6bd8c5180de7a4bc2244
Signed-off-by: Govindraj Raja <govindraj.raja@arm.com>
override_vector_table does adr, followed by an msr ops.
Accidentally was used here for for adr and mrs op.
Signed-off-by: Sona Mathew <SonaRebecca.Mathew@arm.com>
Change-Id: I2d3fda12acd097acabbde9b7dcc376d08419e223
This involves replacing:
* the reset_func with the standard cpu_reset_func_{start,end} to apply
errata automatically
* the <cpu>_errata_report with the errata_report_shim to report errata
automatically
...and for each erratum:
* the prologue with the workaround_<type>_start to do the checks and
framework registration automatically
* the epilogue with the workaround_<type>_end
* the checker function with the check_erratum_<type> to make it more
descriptive
It is important to note that the errata workaround and checking
sequences remain unchanged and preserve their git blame. Testing was
conducted by:
* Building for release with all errata flags enabled and running script
in change 19136 to compare output of objdump for each errata.
* Manual comparison of disassembly of converted functions with non-
converted functions
* Build for debug with all errata enabled and step through ArmDS
at reset to ensure all functions are entered.
Signed-off-by: Harrison Mutai <harrison.mutai@arm.com>
Change-Id: I417539ab292f13a4f0949625d2fef6b7792fbd35
Errata report order is enforced to be in ascending order. To achieve
this with the errata framework this has to be done at the definition
level.
Change-Id: I4a6ed55d48e91ec914b7a591c6d30da5ce5d915d
Signed-off-by: Harrison Mutai <harrison.mutai@arm.com>
* changes:
refactor(cpus): convert Cortex-A15 to use the errata framework
refactor(cpus): convert the Cortex-X3 to use the cpu helpers
refactor(cpus): convert Cortex-X3 to use the errata framework
refactor(cpus): reorder Cortex-X3 errata by ascending order
refactor(cpus): convert the Cortex-A73 to use the cpu helpers
refactor(cpus): convert Cortex-A73 to use the errata framework
refactor(cpus): reorder Cortex-A73 errata by ascending order
refactor(cpus): convert the Cortex-A35 to use the cpu helpers
refactor(cpus): convert Cortex-A35 to use the errata framework
This involves replacing:
* the reset_func with the standard cpu_reset_func_{start,end} to apply
errata automatically
* the <cpu>_errata_report with the errata_report_shim to report errata
automatically
...and for each erratum:
* the prologue with the workaround_<type>_start to do the checks and
framework registration automatically
* the epilogue with the workaround_<type>_end
* the checker function with the check_erratum_<type> to make it more
descriptive
* Manual comparison of disassembly of converted functions with non-
converted functions.
aarch64-none-elf-objdump -D <TF-A with
conversion>/build/../release/bl31/bl31.elf
vs
aarch64-none-elf-objdump -D <TF-A clean
repo>/build/fvp/release/bl31/bl31.elf
* Build for debug with all errata enabled and step through ArmDS
at reset to ensure all functions are entered.
Change-Id: I62e030962edf4e8e8be2c19e7a3176e319468c50
Signed-off-by: Sona Mathew <SonaRebecca.Mathew@arm.com>
Errata report order is enforced to be in ascending order. To achieve
this with the errata framework this has to be done at the definition
level.
Change-Id: I168bf99be0cb0b046d6b641c855f9241991bb0bc
Signed-off-by: Sona Mathew <SonaRebecca.Mathew@arm.com>
This involves replacing:
* the reset_func with the standard cpu_reset_func_{start,end} to apply
errata automatically
* the <cpu>_errata_report with the errata_report_shim to report errata
automatically
...and for each erratum:
* the prologue with the workaround_<type>_start to do the checks and
framework registration automatically
* the epilogue with the workaround_<type>_end
* the checker function with the check_erratum_<type> to make it more
descriptive
It is important to note that the errata workaround and checking
sequences remain unchanged and preserve their git blame. Testing was
conducted by:
* Manual comparison of disassembly of converted functions with non-
converted functions.
aarch64-none-elf-objdump -D <TF-A with
conversion>/build/../release/bl31/bl31.elf
vs
aarch64-none-elf-objdump -D <TF-A clean
repo>/build/fvp/release/bl31/bl31.elf
* Build for release with all errata flags enabled and compare
the disassembly of converted functions with non-converted
functions.
CROSS_COMPILE=aarch64-none-elf- make PLAT=fvp DEBUG=0 \
HW_ASSISTED_COHERENCY=0 BL33=<tf-a-tests>/build/fvp/debug/tftf.bin \
all fip ERRATA_A73_852427=1 \
ERRATA_A73_855423=1 \
WORKAROUND_CVE_2017_5715=1 \
WORKAROUND_CVE_2018_3639=1 \
WORKAROUND_CVE_2022_23960=1
* Build for debug with all errata enabled and step through ArmDS
at reset to ensure all functions are entered.
Change-Id: I63e5b2cc42e1e12daee0b727770cbc19ba729ff7
Signed-off-by: Sona Mathew <SonaRebecca.Mathew@arm.com>
Errata report order is enforced to be in ascending order. To achieve
this with the errata framework this has to be done at the definition
level.
Change-Id: I70b05cc366c3b6d07a63edd88d23a52dd3d019c1
Signed-off-by: Sona Mathew <SonaRebecca.Mathew@arm.com>
This involves replacing:
* the reset_func with the standard cpu_reset_func_{start,end} to apply
errata automatically
* the <cpu>_errata_report with the errata_report_shim to report errata
automatically
...and for each erratum:
* the prologue with the workaround_<type>_start to do the checks and
framework registration automatically
* the epilogue with the workaround_<type>_end
* the checker function with the check_erratum_<type> to make it more
descriptive
It is important to note that the errata workaround and checking
sequences remain unchanged and preserve their git blame. Testing was
conducted by:
* Manual comparison of disassembly of converted functions with non-
converted functions.
aarch64-none-elf-objdump -D <TF-A with
conversion>/build/../release/bl31/bl31.elf
vs
aarch64-none-elf-objdump -D <TF-A with
clean repo>/build/fvp/release/bl31/bl31.elf
* Build for release with all errata flags enabled and ensure the
changes were identical.
CROSS_COMPILE=aarch64-none-elf- make PLAT=fvp \
DEBUG=0 HW_ASSISTED_COHERENCY=0 \
BL33=<tf-a-tests>/build/fvp/debug/tftf.bin \
all fip ERRATA_A35_855472=1
* Build for debug with all errata enabled and step through ArmDS
at reset to ensure all functions are entered.
Change-Id: Ib001e9fc269e60369ccfda0245a3e6247f0d6aaa
Signed-off-by: Sona Mathew <SonaRebecca.Mathew@arm.com>
This involves replacing:
* the reset_func with the standard cpu_reset_func_{start,end} to apply
errata automatically
* the <cpu>_errata_report with the errata_report_shim to report errata
automatically
...and for each erratum:
* the prologue with the workaround_<type>_start to do the checks and
framework registration automatically
* the epilogue with the workaround_<type>_end
* the checker function with the check_erratum_<type> to make it more
descriptive
It is important to note that the errata workaround and checking
sequences remain unchanged and preserve their git blame.
At this point the binary output of all errata was checked with the
script from commit 19136. The reported discrepancies are immaterial.
All errata have been checked that they get invoked.
Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>
Change-Id: Ideb6397f4ac7c3c1d04549a57af43bfa7ef25c1d
