On 7/6/2022 13:21, Sebastian Huber wrote:
On 06/07/2022 20:19, Gedare Bloom wrote:
ok, this changes slightly the order of events in the system state
(idle threads are now created in the preceding state,
SYSTEM_STATE_BEFORE_INITIALIZATION) but without thinking too hard
about it, I think it should be fine.
Yes, it doesn't really matter. Logically, setting the new state after
the work is done is slightly better.
This seems to have broken spextensions01 on AArch64 (both QEMU and on
hardware) and likely elsewhere:
assertion "_Thread_Dispatch_is_enabled() || before_multitasking()"
failed: file "../../../testsuites/sptests/spextensions01/init.c", line
64, function: assert_allocator_protected_thread_context
Kinsey
_______________________________________________
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel
