arm-trusted-firmware/docs
Boyan Karatotev 83ec7e452c perf(amu): greatly simplify AMU context management
The current code is incredibly resilient to updates to the spec and
has worked quite well so far. However, recent implementations expose a
weakness in that this is rather slow. A large part of it is written in
assembly, making it opaque to the compiler for optimisations. The
future proofness requires reading registers that are effectively
`volatile`, making it even harder for the compiler, as well as adding
lots of implicit barriers, making it hard for the microarchitecutre to
optimise as well.

We can make a few assumptions, checked by a few well placed asserts, and
remove a lot of this burden. For a start, at the moment there are 4
group 0 counters with static assignments. Contexting them is a trivial
affair that doesn't need a loop. Similarly, there can only be up to 16
group 1 counters. Contexting them is a bit harder, but we can do with a
single branch with a falling through switch. If/when both of these
change, we have a pair of asserts and the feature detection mechanism to
guard us against pretending that we support something we don't.

We can drop contexting of the offset registers. They are fully
accessible by EL2 and as such are its responsibility to preserve on
powerdown.

Another small thing we can do, is pass the core_pos into the hook.
The caller already knows which core we're running on, we don't need to
call this non-trivial function again.

Finally, knowing this, we don't really need the auxiliary AMUs to be
described by the device tree. Linux doesn't care at the moment, and any
information we need for EL3 can be neatly placed in a simple array.

All of this, combined with lifting the actual saving out of assembly,
reduces the instructions to save the context from 180 to 40, including a
lot fewer branches. The code is also much shorter and easier to read.

Also propagate to aarch32 so that the two don't diverge too much.

Change-Id: Ib62e6e9ba5be7fb9fb8965c8eee148d5598a5361
Signed-off-by: Boyan Karatotev <boyan.karatotev@arm.com>
2025-02-25 08:50:46 +00:00
..
_static/css docs(threat model): add TF-A threat model 2021-04-30 17:59:22 +02:00
about perf(mpmm): greatly simplify MPMM enablement 2025-02-25 08:50:45 +00:00
components perf(amu): greatly simplify AMU context management 2025-02-25 08:50:46 +00:00
design perf(cpus): make reset errata do fewer branches 2025-02-24 09:36:11 +00:00
design_documents refactor(rse): put MHU code in a dedicated file 2025-02-12 10:11:49 +01:00
getting_started perf(amu): greatly simplify AMU context management 2025-02-25 08:50:46 +00:00
perf docs(juno): update PSCI instrumentation data 2024-11-15 13:16:28 +00:00
plat feat(fvp): increase GPT PPS to 1TB 2025-02-11 15:10:49 +00:00
process docs: remove reference to phabricator pages 2024-05-15 14:27:45 +02:00
resources docs: add inital lts doc 2025-01-21 08:40:34 -06:00
security_advisories chore: rename Poseidon to Neoverse V3 2024-03-26 11:27:31 -05:00
threat_model feat(docs): add RSE provided mboot backends to the threat model 2024-11-04 17:25:15 +01:00
tools fix(cot-dt2c): use processed Device Tree source file as input 2024-08-27 12:50:20 +01:00
change-log.md docs(changelog): changelog for v2.12 release 2024-11-19 18:08:58 -06:00
conf.py docs(changelog): changelog for v2.12 release 2024-11-19 18:08:58 -06:00
global_substitutions.txt feat(docs): add DPE to RSE desing doc 2024-11-04 17:28:15 +01:00
glossary.rst docs: add inital lts doc 2025-01-21 08:40:34 -06:00
index.rst feat(fwu): update the URL links for the FWU specification 2024-03-01 14:19:56 +05:30
license.rst feat(lib): introduce Hob creation library 2024-12-06 13:26:31 +00:00
Makefile build: install dependencies before doc build 2024-11-06 15:18:06 +01:00
porting-guide.rst Merge "docs(console): updated console docs" into integration 2025-02-12 15:33:26 +01:00