X-Git-Url: https://codewiz.org/gitweb?a=blobdiff_plain;f=boards%2Fat91sam7s-ek%2Fbenchmark%2Fcontext_switch%2Fmain.c;fp=boards%2Fat91sam7s-ek%2Fbenchmark%2Fcontext_switch%2Fmain.c;h=ef29fb02cd9ff69125d0c93329b9f21f17ce7dcc;hb=21e8827872a940bf5f3c81e437a1c12461b61ea6;hp=30206a69d3eb47f9fd389258d8330d67e58452a5;hpb=7c80c99f018d72e8993c974472ce440dd6b5787a;p=bertos.git diff --git a/boards/at91sam7s-ek/benchmark/context_switch/main.c b/boards/at91sam7s-ek/benchmark/context_switch/main.c index 30206a69..ef29fb02 100644 --- a/boards/at91sam7s-ek/benchmark/context_switch/main.c +++ b/boards/at91sam7s-ek/benchmark/context_switch/main.c @@ -32,13 +32,14 @@ * * \author Daniele Basile * - * \brief Kernel switch context benchmark. + * \brief Kernel context switch benchmark. * - * This benchmark show you the switch context time. This measure is make - * by hp timer that measure accurately the kernel switch context time, and print - * it on serial debug. If you want not use the hp timer you can measure this - * time using one oscilloscope, connected to one board led. To use this - * mode you should disable hp timer and implement hw_led.h low level module. + * This benchmark measures the overhead (in µs) required to deliver a + * signal and switch from a low-priority process to a high-priority process. + * + * The measure is made by the high-precision timer if supported by the + * architecture. Alternatively, the time can also be measured using an external + * oscilloscope connected to the board's status LED. */ #include