GCC Cortex-A9 with hard FP Timing Benchmarks
Target Platform: ti.platforms.evmAM437X
Tool Chain Version: 9.2.1
BIOS Version: bios_6_83_00_16_eng
XDCTools Version: xdctools_3_61_04_30_core_eng
Benchmark | Cycles |
---|---|
Interrupt Latency | 321 |
Hwi_restore() | 8 |
Hwi_disable() | 4 |
Hwi dispatcher prolog | 190 |
Hwi dispatcher epilog | 162 |
Hwi dispatcher | 352 |
Hardware Interrupt to Blocked Task | 561 |
Hardware Interrupt to Software Interrupt | 373 |
Swi_enable() | 61 |
Swi_disable() | 12 |
Post Software Interrupt Again | 24 |
Post Software Interrupt without Context Switch | 81 |
Post Software Interrupt with Context Switch | 151 |
Create a New Task without Context Switch | 1398 |
Set a Task Priority without a Context Switch | 103 |
Task_yield() | 251 |
Post Semaphore No Waiting Task | 81 |
Post Semaphore No Task Switch | 172 |
Post Semaphore with Task Switch | 266 |
Pend on Semaphore No Context Switch | 40 |
Pend on Semaphore with Task Switch | 275 |
Clock_getTicks() | 15 |
POSIX Create a New Task without Context Switch | 2550 |
POSIX Set a Task Priority without a Context Switch | 88 |
POSIX Post Semaphore No Waiting Task | 84 |
POSIX Post Semaphore No Task Switch | 172 |
POSIX Post Semaphore with Task Switch | 300 |
POSIX Pend on Semaphore No Context Switch | 43 |
POSIX Pend on Semaphore with Task Switch | 290 |
As the timer used to run these benchmarks is run at 1/2 the CPU frequency, an deviation of +-2 cycles is to be expected.
See the SYS/BIOS User’s Guide for more information regarding how interrupt latency is calculated.