Skip to content

Timer spec#476

Merged
bokdeuk-jeong merged 7 commits intomainfrom
timer-spec
Dec 2, 2025
Merged

Timer spec#476
bokdeuk-jeong merged 7 commits intomainfrom
timer-spec

Conversation

@bokdeuk-jeong
Copy link
Collaborator

@bokdeuk-jeong bokdeuk-jeong commented Nov 26, 2025

  1. Add feature 'ns_state_save' to rmm, which is used when the normal world like the acs test doesn't saved/restore its own state before/after RMM calls.
  2. Add code relaying wfi/wfe trap from a realm to the ns host and configuring wfi/wfe traps set by the ns host.
  3. Implement timer spec additionally.
  4. Fix multi-core issues. Removed code printing log messages in the plat directory. Don't access global data struct at the very early stage of core boot-up process with MMU off (==dcache off).
    It now passes the muti-core test in acs, but the test case causes an instruction abort within the normal world. Although, it has no relation with the mp test itself, I left the test case in the skipped-tests.txt to continue performing remaining tests.

To pass gic_timer_rel1_trig test suite in the acs test,
add the timer state management for the non-sec world.

Signed-off-by: Bokdeuk Jeong <[email protected]>
Signed-off-by: Bokdeuk Jeong <[email protected]>
@bokdeuk-jeong bokdeuk-jeong force-pushed the timer-spec branch 3 times, most recently from da043d8 to 6100a81 Compare November 26, 2025 06:49
Do not print messages until MMU is turned on, except on cpu0.
Cores other than cpu0 at this point are still in its mmu
off state with d-cache disabled and i-cache enabled.
This may cause incosistencies between cpus with mmu enabled
and cpus with mmu disabled.
Logging involves buffer allocation and its global data struct
for heap (linked_list) could be corrupted due to the reason above.

Signed-off-by: Bokdeuk Jeong <[email protected]>
Update acs test count but leave cmd_multithread_realm_mp test excluded.

cmd_mutithread_realm_mp test passes now. However, an Inst abort took place
at the end of the mp test within the host for unknown reason.

Signed-off-by: Bokdeuk Jeong <[email protected]>
@bokdeuk-jeong bokdeuk-jeong force-pushed the timer-spec branch 2 times, most recently from aa81418 to 81f1438 Compare November 28, 2025 02:26
I<VRWGS>:
On REC entry, for both the EL1 Virtual Timer and the EL1 Physical
Timer, if the EL1 timer asserts its output in the state described
in the REC exit structure from the previous REC exit then the RMM
masks the hardware timer signal before returning to the Realm.

Signed-off-by: Bokdeuk Jeong <[email protected]>
@bokdeuk-jeong bokdeuk-jeong marked this pull request as ready for review November 28, 2025 05:10
@bokdeuk-jeong bokdeuk-jeong merged commit 095b7d1 into main Dec 2, 2025
8 checks passed
@bokdeuk-jeong bokdeuk-jeong deleted the timer-spec branch December 2, 2025 10:28
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant