MPMM is a core-specific microarchitectural feature. It has been present
in every Arm core since the Cortex-A510 and has been implemented in
exactly the same way. Despite that, it is enabled more like an
architectural feature with a top level enable flag. This utilised the
identical implementation.
This duality has left MPMM in an awkward place, where its enablement
should be generic, like an architectural feature, but since it is not,
it should also be core-specific if it ever changes. One choice to do
this has been through the device tree.
This has worked just fine so far, however, recent implementations expose
a weakness in that this is rather slow - the device tree has to be read,
there's a long call stack of functions with many branches, and system
registers are read. In the hot path of PSCI CPU powerdown, this has a
significant and measurable impact. Besides it being a rather large
amount of code that is difficult to understand.
Since MPMM is a microarchitectural feature, its correct placement is in
the reset function. The essence of the current enablement is to write
CPUPPMCR_EL3.MPMM_EN if CPUPPMCR_EL3.MPMMPINCTL == 0. Replacing the C
enablement with an assembly macro in each CPU's reset function achieves
the same effect with just a single close branch and a grand total of 6
instructions (versus the old 2 branches and 32 instructions).
Having done this, the device tree entry becomes redundant. Should a core
that doesn't support MPMM arise, this can cleanly be handled in the
reset function. As such, the whole ENABLE_MPMM_FCONF and platform hooks
mechanisms become obsolete and are removed.
Change-Id: I1d0475b21a1625bb3519f513ba109284f973ffdf
Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>
Errata application is painful for performance. For a start, it's done
when the core has just come out of reset, which means branch predictors
and caches will be empty so a branch to a workaround function must be
fetched from memory and that round trip is very slow. Then it also runs
with the I-cache off, which means that the loop to iterate over the
workarounds must also be fetched from memory on each iteration.
We can remove both branches. First, we can simply apply every erratum
directly instead of defining a workaround function and jumping to it.
Currently, no errata that need to be applied at both reset and runtime,
with the same workaround function, exist. If the need arose in future,
this should be achievable with a reset + runtime wrapper combo.
Then, we can construct a function that applies each erratum linearly
instead of looping over the list. If this function is part of the reset
function, then the only "far" branches at reset will be for the checker
functions. Importantly, this mitigates the slowdown even when an erratum
is disabled.
The result is ~50% speedup on N1SDP and ~20% on AArch64 Juno on wakeup
from PSCI calls that end in powerdown. This is roughly back to the
baseline of v2.9, before the errata framework regressed on performance
(or a little better). It is important to note that there are other
slowdowns since then that remain unknown.
Change-Id: Ie4d5288a331b11fd648e5c4a0b652b74160b07b9
Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>
Cortex-A715 erratum 3699560 that applies to all revisions <= r1p3
and is still Open.
The workaround is for EL3 software that performs context save/restore
on a change of Security state to use a value of SCR_EL3.NS when
accessing ICH_VMCR_EL2 that reflects the Security state that owns the
data being saved or restored.
SDEN documentation:
https://developer.arm.com/documentation/SDEN-2148827/latest/
Change-Id: I183aa921b4b6f715d64eb6b70809de2566017d31
Signed-off-by: Govindraj Raja <govindraj.raja@arm.com>
Errata printing is done directly via generic_errata_report.
This commit removes the unused \_cpu\()_errata_report
functions for all cores, and removes errata_func from cpu_ops.
Change-Id: I04fefbde5f0ff63b1f1cd17c864557a14070d68c
Signed-off-by: Ryan Everett <ryan.everett@arm.com>
Cortex-A715 erratum 2728106 is a Cat B(rare) erratum that is present
in revision r0p0, r1p0 and r1p1. It is fixed in r1p2.
The workaround is to execute an implementation specific sequence in
the CPU.
SDEN documentation:
https://developer.arm.com/documentation/SDEN2148827/latest
Change-Id: Ic825f9942e7eb13893fdbb44a2090b897758cbc4
Signed-off-by: Bipin Ravi <bipin.ravi@arm.com>
Erratum 2413290 is a Cat B erratum that is present only
in revision r0p1 and is fixed in r1p1.
The initial implementation did not consider that this
fix is to be applied only when SPE (Statistical Profiling
Extension) is implemented and enabled. This patch applies
the fix by adding a check for ENABLE_SPE_FOR_NS.
Change-Id: I87b2175b89d6fb168c77e6ab233c90ca056791a1
Signed-off-by: Sona Mathew <sonarebecca.mathew@arm.com>
Cortex-A715 erratum 2561034 mitigation needs to be applied
during reset. This patch fixes the current macro usage from runtime
to reset for both start and end macros.
Change-Id: I4f115bbb27c57f16cada2a7eb314af8380f93cb4
Signed-off-by: Bipin Ravi <biprav01@u203721.austin.arm.com>
Cortex-A715 erratum 2413290 is a Cat B erratum that is present
only in revision r1p0 and is fixed in r1p1. The errata is only
present when SPE(Statistical Profiling Extension) is enabled.
The workaround is to set bits[58:57] of the CPUACTLR_EL1 to 'b11
when SPE is enabled, ENABLE_SPE_FOR_NS=1.
SDEN documentation:
https://developer.arm.com/documentation/SDEN2148827/latest
Change-Id: Iaeb258c8b0a92e93d70b7dad6ba59d1056aeb135
Signed-off-by: Sona Mathew <sonarebecca.mathew@arm.com>
Cortex-A715 erratum 2344187 is a Cat B erratum that applies to r0p0,
r1p0 and is fixed in r1p1. The workaround is to set GCR_EL1.RRND to
0b1, and apply an implementation specific patch sequence.
SDEN: https://developer.arm.com/documentation/SDEN2148827/latest
Change-Id: I78ea39a91254765c964bff89f771af33b23f29c1
Signed-off-by: Harrison Mutai <harrison.mutai@arm.com>
Cortex-A715 erratum 2331818 is a cat B erratum that applies to
revisions r0p0 and r1p0 and is fixed in r1p1. The workaround is to
set bit[20] of CPUACTLR2_EL1. Setting this bit is expected to have
a negligible performance impact.
SDEN can be found here:
https://developer.arm.com/documentation/SDEN2148827/latest
Change-Id: If3b1ed78b145ab6515cdd41135314350ed556381
Signed-off-by: Bipin Ravi <biprav01@u203721.austin.arm.com>
Cortex-A715 erratum 2420947 is a cat B erratum that applies only
to revision r1p0 and is fixed in r1p1. The workaround is to set
bit[33] of CPUACTLR2_EL1. This will prevent store and store-release
to merge inside the write buffer, and it is not expected to have
much performance impacts.
SDEN can be found here:
https://developer.arm.com/documentation/SDEN2148827/latest
Change-Id: I01a71b878cd958e742ff8357f8cdfbfc5625de47
Signed-off-by: Bipin Ravi <biprav01@u203721.austin.arm.com>
Cortex-A715 erratum 2429384 is a cat B erratum that applies to
revision r1p0 and is fixed in r1p1. The workaround is to set
bit[27] of CPUACTLR2_EL1. There is no workaround for revision
r0p0.
SDEN can be found here:
https://developer.arm.com/documentation/SDEN2148827/latest
Change-Id: I3cdb1b71567542174759f6946e9c81f77d0d993d
Signed-off-by: Bipin Ravi <biprav01@u203721.austin.arm.com>
Cortex-A715 erratum 2561034 is a Cat B erratum that applies to
revision r1p0 and is fixed in r1p1.
The workaround is to set bit[26] in CPUACTLR2_EL1. Setting this
bit is not expected to have a significant performance impact.
SDEN documentation:
https://developer.arm.com/documentation/SDEN2148827/latest
Change-Id: I377f250a2994b6ced3ac7d93f947af6ceb690d49
Signed-off-by: Bipin Ravi <bipin.ravi@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:
* 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.
Change-Id: Ib63b6310997d523fa8bd7f867e53fedec66f1e06
Signed-off-by: Harrison Mutai <harrison.mutai@arm.com>
* ASM files are renamed to have public IP names in their filename.
* updated other files to include ASM filename changes.
Signed-off-by: Rupinderjit Singh <rupinderjit.singh@arm.com>
Change-Id: Ie899c512b11fd7c4312e3a808bb6b9d2376cdb8c
2022-08-31 18:31:29 +01:00
Renamed from lib/cpus/aarch64/cortex_makalu.S (Browse further)
