Issue created by Wayne Thornton:
https://gitlab.rtems.org/rtems/rtos/rtems/-/work_items/5646
## Description:
We have introduced a new Super Core architecture hook, `_CPU_Spin_wait()`, to
mitigate aggressive speculative execution penalties, reduce thermal load, and
lower bus contention during SMP spin-wait polling loops. The reference template
is currently documented in `cpukit/score/cpu/no_cpu/include/rtems/score/cpu.h`
and natively implemented for `x86_64` upon merging of rtems/rtos/rtems!1256.
The powerpc architecture port needs to override the default fallback by
defining this interface in `cpukit/score/cpu/powerpc/include/rtems/score/cpu.h`.
Suggested Implementation:
On SMT-capable PowerPC cores (such as e6500 or POWER8/POWER9), thread priority
hints can be used to yield processor execution cycles to sibling hardware
threads (e.g., setting low thread priority via `or 1,1,1` or `or 27,27,27`,
followed by restoring medium priority `or 2,2,2`). For generic PowerPC builds
where SMT hints are unsupported, explicitly apply the compiler barrier:
```
static inline void _CPU_Spin_wait( void )
{
/* Add architecture-specific thread priority yield hints if supported by
target CPU model */
RTEMS_COMPILER_MEMORY_BARRIER();
}
#define _CPU_Spin_wait _CPU_Spin_wait
```
## Verification:
Validate that the implementation compiles cleanly across standard PowerPC SMP
targets and passes the standalone SMP spin-wait test suite:
`testsuites/smptests/smpspinwait01.` Once defined, the preprocessor will
automatically override the temporary test-level fallback in `init.c`.
<!-- Pre-set options
- milestone
-->
--
View it on GitLab: https://gitlab.rtems.org/rtems/rtos/rtems/-/work_items/5646
You're receiving this email because of your account on gitlab.rtems.org.
Unsubscribe from this thread:
https://gitlab.rtems.org/-/sent_notifications/4-c7x3o07j3mz96wlm5r88oe4fm-1d/unsubscribe
| Manage all notifications: https://gitlab.rtems.org/-/profile/notifications |
Help: https://gitlab.rtems.org/help
_______________________________________________
bugs mailing list
[email protected]
http://lists.rtems.org/mailman/listinfo/bugs
