9.1. PikaPython configuration manual

9.1.1. When to configure

PikaPython itself is configuration-free, so usually you don’t need to know this part.

You can consider configuring PikaPython when you have the following requirements:

  • faster speed

  • Smaller memory footprint

  • Replace dependencies (libc, pinrtf, etc.)

  • Replace memory management algorithm ( malloc )

  • Safer interruption protection

9.1.2. Optimization

[Note]: For optimized configuration, the kernel version needs to be at least v1.5.4

Similar to GCC, PikaPython also provides different optimization modes. The currently available optimization modes are:

  • PIKA_OPTIMIZE_SIZE volume mode minimizes running memory

  • PIKA_OPTIMIZE_SPEED performance mode maximizes running speed Enable user configuration

User configuration is not enabled by default. The way to enable user configuration is to add compile-time macro definition PIKA_CONFIG_ENABLE. Then create the pika_config.h header file.

It should be noted that the PIKA_CONFIG_ENABLE macro should be added to the compile options, such as keil:

_images/160849244-40fe7fa8-0e93-4791-8f14-bc044bbd0d59.png Configuration items

Available configuration items and default configuration are in the pika_config_valid.h header file.


Intercept the important part for explanation:

    /* optimize options */
        #define PIKA_OPTIMIZE_SIZE 0
        #define PIKA_OPTIMIZE_SPEED 1

    /* syntax support level */

    /* use user config */
        #include "pika_config.h"

    /* default optimize */
    #ifndef PIKA_OPTIMIZE

    /* default syntax support level */

    /* default configuration  */
        #define PIKA_STACK_BUFF_SIZE 256

default configuration is the default value of the configuration item. When the PIKA_CONFIG_ENABLE macro is defined, pika_config_valid.h will import pika_config.h, so User can override the above default configuration in pika_config.h.

For example, if you want to increase the runtime stack of the PikaPython virtual machine, you can write in pika_config.h


As can be seen from pika_config_valid.h, the default optimization option of PikaPython PIKA_OPTIMIZE is PIKA_OPTIMIZE_SIZE, if you need to switch to speed optimization, you can write in pika_config.h



9.1.3. Dependency configuration

PikaPython can be configured by creating pika_config.c, rewriting the weak functions in PikaPlagform.h ‘s dependencies.

/* interrupt config */
void __platform_enable_irq_handle(void);
void __platform_disable_irq_handle(void);

/* printf family config */
#ifndef __platform_printf
void __platform_printf(char* fmt, ...);
int __platform_sprintf(char* buff, char* fmt, ...);
int __platform_vsprintf(char* buff, char* fmt, va_list args);
int __platform_vsnprintf(char* buff,
                         size_t size,
                         const char* fmt,
                         va_list args);

/* libc config */
void* __platform_malloc(size_t size);
void __platform_free(void* ptr);
void* __platform_memset(void* mem, int ch, size_t size);
void* __platform_memcpy(void* dir, const void* src, size_t size);

/* pika memory pool config */
void __platform_wait(void);
uint8_t __is_locked_pikaMemory(void);

/* support shell */
char __platform_getchar(void);

/* file API */
FILE* __platform_fopen(const char* filename, const char* modes);
int __platform_fclose(FILE* stream);
size_t __platform_fwrite(const void* ptr, size_t size, size_t n, FILE* stream);

/* error */
void __platform_error_handle(void); Configuration items:

  • Interrupt Protection - Provides an interrupt master switch to protect PikaPython memory safety

  • libC - select the implementation of libC

  • Memory management - replace malloc free memory management algorithm Sample code: