comfilepi:improving_real-time_performance:index
Differences
This shows you the differences between two versions of the page.
| Both sides previous revision Previous revision Next revision | Previous revision | ||
|
comfilepi:improving_real-time_performance:index [2023/05/12 11:08] COMFILE Technology [Setting the CPU Governor to ''performance''] |
comfilepi:improving_real-time_performance:index [2024/03/13 15:42] (current) COMFILE Technology [Schedule the Program as Real-Time with High Priority] |
||
|---|---|---|---|
| Line 30: | Line 30: | ||
| ===== Isolate a Program to One of the 4 CPU Cores ===== | ===== Isolate a Program to One of the 4 CPU Cores ===== | ||
| - | First, add the ''isolcpus=3'' kernel parameter to the ///boot/cmdline.txt// file. This will prevent Linux from allocating the 4th core to any processes. Reboot for the change to take effect. See [[https://www.kernel.org/doc/html/latest/admin-guide/kernel-parameters.html?highlight=isolcpus|the Linux kernel reference for isolcpus]] for more information. | + | First, add ''isolcpus=nohz,domain,managed_irq,3 irqaffinity=0-2 nohz_full=3 rcu_nocbs=3'' to the ///boot/cmdline.txt// file. This will prevent Linux from allocating the 4th core to most process, interrupt handlers, and kernel threads. Reboot for the change to take effect. See [[https://www.kernel.org/doc/html/latest/admin-guide/kernel-parameters.html|the Linux kernel documentation]] for more information on these kernel arguments. |
| Then use ''taskset'' to have Linux run a program, in this case a program called //test//, on the 4th core. The program will have that core all to itself. See [[https://www.man7.org/linux/man-pages/man1/taskset.1.html|the taskset man page]] for more information. | Then use ''taskset'' to have Linux run a program, in this case a program called //test//, on the 4th core. The program will have that core all to itself. See [[https://www.man7.org/linux/man-pages/man1/taskset.1.html|the taskset man page]] for more information. | ||
| Line 36: | Line 36: | ||
| sudo taskset -cp 3 `pidof test` | sudo taskset -cp 3 `pidof test` | ||
| </code> | </code> | ||
| + | |||
| + | Core isolation can also be achieved using //cpusets//. See the following resources for more information: | ||
| + | * [[https://docs.kernel.org/admin-guide/cgroup-v1/cpusets.html|Linux kernel documentation on cpusets]] | ||
| + | * [[https://github.com/lpechacek/cpuset/blob/master/doc/tutorial.txt|cpuset utility package documentation]] | ||
| ===== Schedule the Program as Real-Time with High Priority ===== | ===== Schedule the Program as Real-Time with High Priority ===== | ||
| Line 42: | Line 46: | ||
| <code> | <code> | ||
| - | sudo chrt -f -p `pidof test` | + | sudo chrt -f -p 99 `pidof test` |
| </code> | </code> | ||
| Line 49: | Line 53: | ||
| - | ===== Disable Realtime Throttling ===== | + | ===== Disable Realtime Throttling (Not Recommended) ===== |
| - | When using FIFO scheduling, it may help be necessary to disable [[https://lwn.net/Articles/296419/|realtime throttling]]. | + | When using FIFO scheduling, it may be necessary to disable [[https://lwn.net/Articles/296419/|realtime throttling]]. |
| <code> | <code> | ||
| Line 57: | Line 61: | ||
| </code> | </code> | ||
| - | However, this is a blunt instrument because it disables realtime throttling on all cores. Instead it may be best to introduce a ''sleep'' along with a spin-wait for greater precision. For more information, refer to [[https://stackoverflow.com/a/75122036/539972|this description]]. | + | However, this is a blunt instrument because it disables realtime throttling on all cores. Instead, to implement precise timing, it may be best to introduce a ''sleep'' along with a spin-wait for greater precision. For more information, refer to [[https://stackoverflow.com/a/75122036/539972|this description]]. |
| ===== Demonstration, Measurement, and Results ===== | ===== Demonstration, Measurement, and Results ===== | ||
| Line 63: | Line 67: | ||
| ==== Test Program ==== | ==== Test Program ==== | ||
| - | In this test, a program is created to output a 100µs alternating pulse (500kHz) on one of the ComfilePI's GPIO pins. The program uses the [[http://www.airspayce.com/mikem/bcm2835/|BCM2835]] library. | + | In this test, a program is created to output a 100µs alternating pulse (500kHz) on one of the ComfilePI's GPIO pins. The program uses the [[https://git.kernel.org/pub/scm/libs/libgpiod/libgpiod.git/|gpiod]] library. |
| <code C++> | <code C++> | ||
| + | #include <gpiod.h> | ||
| #include <cstdio> | #include <cstdio> | ||
| - | #include <bcm2835.h> | + | #include <chrono> |
| + | #include <unistd.h> | ||
| - | #define PIN 4 | + | using namespace std::chrono; |
| - | #define DELAY 100 | + | |
| - | int main() | + | void delay(uint32_t us) |
| { | { | ||
| - | if (!bcm2835_init()) | + | auto t1 = steady_clock::now(); |
| - | { | + | auto t2 = t1 + microseconds(us); |
| - | printf("bcm2835_init() failed\n"); | + | this_thread::sleep_for(10us); |
| - | exit(1); | + | while (steady_clock::now() < t2); |
| - | } | + | } |
| - | bcm2835_gpio_fsel(PIN, BCM2835_GPIO_FSEL_OUTP); | + | int main(int argc, char **argv) |
| + | { | ||
| + | const char *chipname = "gpiochip0"; | ||
| + | struct gpiod_chip *chip; | ||
| + | struct gpiod_line *line16; | ||
| - | for(;;) | + | // Open GPIO16 |
| + | chip = gpiod_chip_open_by_name(chipname); | ||
| + | line16 = gpiod_chip_get_line(chip, 16); | ||
| + | gpiod_line_request_output(line16, "realtime", 0); | ||
| + | |||
| + | // Toggle GPIO every 100us | ||
| + | while (true) | ||
| { | { | ||
| - | bcm2835_gpio_write(PIN, HIGH); | + | gpiod_line_set_value(line16, true); |
| - | bcm2835_st_delay(bcm2835_st_read(), DELAY); | + | delay(50); |
| - | bcm2835_gpio_write(PIN, LOW); | + | gpiod_line_set_value(line16, false); |
| - | bcm2835_st_delay(bcm2835_st_read(), DELAY); | + | delay(50); |
| } | } | ||
| + | |||
| + | // Release lines and chip | ||
| + | gpiod_line_release(line16); | ||
| + | gpiod_chip_close(chip); | ||
| + | |||
| + | return 0; | ||
| } | } | ||
| </code> | </code> | ||
| Line 94: | Line 115: | ||
| Compile with... | Compile with... | ||
| <code> | <code> | ||
| - | gcc -O3 test.cpp -lbcm2835 -o test | + | # Install the gpiod library if necessary |
| + | # sudo apt install libgpiod-dev | ||
| + | |||
| + | # Compile the program | ||
| + | g++ -O3 test.cpp -lgpiod -o test | ||
| </code> | </code> | ||
| Line 128: | Line 153: | ||
| {{ :comfilepi:improving_real-time_performance:jitter.png?900 }} | {{ :comfilepi:improving_real-time_performance:jitter.png?900 }} | ||
| - | As can be seen in the chart above, the ordinary kernel, without any CPU isolation or real-time scheduling, operates with excessive jitter and would likely not be suitable for any real-time application. However, by taking a few measures to optimize a process for real-time performance, it is possible to reduce jitter potentially below 50µs which may be acceptable for some real-time use cases. | + | As can be seen in the chart above, the ordinary kernel, without any CPU isolation or real-time scheduling, operates with excessive jitter and would likely not be suitable for any real-time application. However, by taking a few measures to optimize a process for real-time performance, it is possible to reduce jitter potentially below 50µs which may be acceptable for some real-time use cases. Without the intentional disturbance introduced by running the web browser, the performance is better. |
| Although the real-time kernel can reduce jitter, it trades off overall performance throughput. Using the real-time kernel may cause the system to boot slower and run some applications with a mild performance penalty, but it will provide more deterministic behavior. | Although the real-time kernel can reduce jitter, it trades off overall performance throughput. Using the real-time kernel may cause the system to boot slower and run some applications with a mild performance penalty, but it will provide more deterministic behavior. | ||
comfilepi/improving_real-time_performance/index.1683857307.txt.gz · Last modified: 2023/05/12 11:08 by COMFILE Technology