alex/esp-idf
alex/esp-idf
1
0
Fork 0
mirror of https://github.com/espressif/esp-idf.git synced 2024-10-05 20:47:46 -04:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity
4ce00a4829
esp-idf/components/freertos/portmux_impl.h

113 lines
4.6 KiB
C
Raw Normal View History

freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
/*
Copyright (C) 2016-2017 Espressif Shanghai PTE LTD
Copyright (C) 2015 Real Time Engineers Ltd.
All rights reserved
FreeRTOS is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License (version 2) as published by the
Free Software Foundation >>!AND MODIFIED BY!<< the FreeRTOS exception.
***************************************************************************
>>! NOTE: The modification to the GPL is included to allow you to !<<
>>! distribute a combined work that includes FreeRTOS without being !<<
>>! obliged to provide the source code for proprietary components !<<
>>! outside of the FreeRTOS kernel. !<<
***************************************************************************
FreeRTOS is distributed in the hope that it will be useful, but WITHOUT ANY
WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
FOR A PARTICULAR PURPOSE. Full license text is available on the following
link: http://www.freertos.org/a00114.html
*/
/* This header exists for performance reasons, in order to inline the
implementation of vPortCPUAcquireMutexIntsDisabled and
vPortCPUReleaseMutexIntsDisabled into the
portmux: Add vPortCPUAcquireMutexTimeout() function Refactor app_trace locking to use this function.
2017-07-20 02:34:45 -04:00
vTaskEnterCritical/vTaskExitCritical functions in task.c as well as the
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
vPortCPUAcquireMutex/vPortCPUReleaseMutex implementations.
portmux: Add vPortCPUAcquireMutexTimeout() function Refactor app_trace locking to use this function.
2017-07-20 02:34:45 -04:00
Normally this kind of performance hack is over the top, but
vTaskEnterCritical/vTaskExitCritical is called a great
deal by FreeRTOS internals.
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
It should be #included by freertos port.c or tasks.c, in esp-idf.
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
The way it works is that it essentially uses portmux_impl.inc.h as a
global: move the soc component out of the common list This MR removes the common dependency from every IDF components to the SOC component. Currently, in the ``idf_functions.cmake`` script, we include the header path of SOC component by default for all components. But for better code organization (or maybe also benifits to the compiling speed), we may remove the dependency to SOC components for most components except the driver and kernel related components. In CMAKE, we have two kinds of header visibilities (set by include path visibility): (Assume component A --(depends on)--> B, B is the current component) 1. public (``COMPONENT_ADD_INCLUDEDIRS``): means this path is visible to other depending components (A) (visible to A and B) 2. private (``COMPONENT_PRIV_INCLUDEDIRS``): means this path is only visible to source files inside the component (visible to B only) and we have two kinds of depending ways: (Assume component A --(depends on)--> B --(depends on)--> C, B is the current component) 1. public (```COMPONENT_REQUIRES```): means B can access to public include path of C. All other components rely on you (A) will also be available for the public headers. (visible to A, B) 2. private (``COMPONENT_PRIV_REQUIRES``): means B can access to public include path of C, but don't propagate this relation to other components (A). (visible to B) 1. remove the common requirement in ``idf_functions.cmake``, this makes the SOC components invisible to all other components by default. 2. if a component (for example, DRIVER) really needs the dependency to SOC, add a private dependency to SOC for it. 3. some other components that don't really depends on the SOC may still meet some errors saying "can't find header soc/...", this is because it's depended component (DRIVER) incorrectly include the header of SOC in its public headers. Moving all this kind of #include into source files, or private headers 4. Fix the include requirements for some file which miss sufficient #include directives. (Previously they include some headers by the long long long header include link) This is a breaking change. Previous code may depends on the long include chain. You may need to include the following headers for some files after this commit: - soc/soc.h - soc/soc_memory_layout.h - driver/gpio.h - esp_sleep.h The major broken include chain includes: 1. esp_system.h no longer includes esp_sleep.h. The latter includes driver/gpio.h and driver/touch_pad.h. 2. ets_sys.h no longer includes soc/soc.h 3. freertos/portmacro.h no longer includes soc/soc_memory_layout.h some peripheral headers no longer includes their hw related headers, e.g. rom/gpio.h no longer includes soc/gpio_pins.h and soc/gpio_reg.h BREAKING CHANGE
2019-04-03 01:17:38 -04:00
generator template of sorts. When no external memory is used, this
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
template is only used to generate the vPortCPUAcquireMutexIntsDisabledInternal
and vPortCPUReleaseMutexIntsDisabledInternal functions, which use S32C1 to
do an atomic compare & swap. When external memory is used the functions
vPortCPUAcquireMutexIntsDisabledExtram and vPortCPUReleaseMutexIntsDisabledExtram
are also generated, which use uxPortCompareSetExtram to fake the S32C1 instruction.
global: move the soc component out of the common list This MR removes the common dependency from every IDF components to the SOC component. Currently, in the ``idf_functions.cmake`` script, we include the header path of SOC component by default for all components. But for better code organization (or maybe also benifits to the compiling speed), we may remove the dependency to SOC components for most components except the driver and kernel related components. In CMAKE, we have two kinds of header visibilities (set by include path visibility): (Assume component A --(depends on)--> B, B is the current component) 1. public (``COMPONENT_ADD_INCLUDEDIRS``): means this path is visible to other depending components (A) (visible to A and B) 2. private (``COMPONENT_PRIV_INCLUDEDIRS``): means this path is only visible to source files inside the component (visible to B only) and we have two kinds of depending ways: (Assume component A --(depends on)--> B --(depends on)--> C, B is the current component) 1. public (```COMPONENT_REQUIRES```): means B can access to public include path of C. All other components rely on you (A) will also be available for the public headers. (visible to A, B) 2. private (``COMPONENT_PRIV_REQUIRES``): means B can access to public include path of C, but don't propagate this relation to other components (A). (visible to B) 1. remove the common requirement in ``idf_functions.cmake``, this makes the SOC components invisible to all other components by default. 2. if a component (for example, DRIVER) really needs the dependency to SOC, add a private dependency to SOC for it. 3. some other components that don't really depends on the SOC may still meet some errors saying "can't find header soc/...", this is because it's depended component (DRIVER) incorrectly include the header of SOC in its public headers. Moving all this kind of #include into source files, or private headers 4. Fix the include requirements for some file which miss sufficient #include directives. (Previously they include some headers by the long long long header include link) This is a breaking change. Previous code may depends on the long include chain. You may need to include the following headers for some files after this commit: - soc/soc.h - soc/soc_memory_layout.h - driver/gpio.h - esp_sleep.h The major broken include chain includes: 1. esp_system.h no longer includes esp_sleep.h. The latter includes driver/gpio.h and driver/touch_pad.h. 2. ets_sys.h no longer includes soc/soc.h 3. freertos/portmacro.h no longer includes soc/soc_memory_layout.h some peripheral headers no longer includes their hw related headers, e.g. rom/gpio.h no longer includes soc/gpio_pins.h and soc/gpio_reg.h BREAKING CHANGE
2019-04-03 01:17:38 -04:00
The wrapper functions vPortCPUAcquireMutexIntsDisabled and
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
vPortCPUReleaseMutexIntsDisabled will then use the appropriate function to do the
actual lock/unlock.
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
*/
#include "soc/cpu.h"
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
#include "portable.h"
global: move the soc component out of the common list This MR removes the common dependency from every IDF components to the SOC component. Currently, in the ``idf_functions.cmake`` script, we include the header path of SOC component by default for all components. But for better code organization (or maybe also benifits to the compiling speed), we may remove the dependency to SOC components for most components except the driver and kernel related components. In CMAKE, we have two kinds of header visibilities (set by include path visibility): (Assume component A --(depends on)--> B, B is the current component) 1. public (``COMPONENT_ADD_INCLUDEDIRS``): means this path is visible to other depending components (A) (visible to A and B) 2. private (``COMPONENT_PRIV_INCLUDEDIRS``): means this path is only visible to source files inside the component (visible to B only) and we have two kinds of depending ways: (Assume component A --(depends on)--> B --(depends on)--> C, B is the current component) 1. public (```COMPONENT_REQUIRES```): means B can access to public include path of C. All other components rely on you (A) will also be available for the public headers. (visible to A, B) 2. private (``COMPONENT_PRIV_REQUIRES``): means B can access to public include path of C, but don't propagate this relation to other components (A). (visible to B) 1. remove the common requirement in ``idf_functions.cmake``, this makes the SOC components invisible to all other components by default. 2. if a component (for example, DRIVER) really needs the dependency to SOC, add a private dependency to SOC for it. 3. some other components that don't really depends on the SOC may still meet some errors saying "can't find header soc/...", this is because it's depended component (DRIVER) incorrectly include the header of SOC in its public headers. Moving all this kind of #include into source files, or private headers 4. Fix the include requirements for some file which miss sufficient #include directives. (Previously they include some headers by the long long long header include link) This is a breaking change. Previous code may depends on the long include chain. You may need to include the following headers for some files after this commit: - soc/soc.h - soc/soc_memory_layout.h - driver/gpio.h - esp_sleep.h The major broken include chain includes: 1. esp_system.h no longer includes esp_sleep.h. The latter includes driver/gpio.h and driver/touch_pad.h. 2. ets_sys.h no longer includes soc/soc.h 3. freertos/portmacro.h no longer includes soc/soc_memory_layout.h some peripheral headers no longer includes their hw related headers, e.g. rom/gpio.h no longer includes soc/gpio_pins.h and soc/gpio_reg.h BREAKING CHANGE
2019-04-03 01:17:38 -04:00
#include "soc/soc_memory_layout.h"
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
/* XOR one core ID with this value to get the other core ID */
#define CORE_ID_XOR_SWAP (CORE_ID_PRO ^ CORE_ID_APP)
portmux: Add vPortCPUAcquireMutexTimeout() function Refactor app_trace locking to use this function.
2017-07-20 02:34:45 -04:00
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
//Define the mux routines for use with muxes in internal RAM
#define PORTMUX_AQUIRE_MUX_FN_NAME vPortCPUAcquireMutexIntsDisabledInternal
#define PORTMUX_RELEASE_MUX_FN_NAME vPortCPUReleaseMutexIntsDisabledInternal
#define PORTMUX_COMPARE_SET_FN_NAME uxPortCompareSet
#include "portmux_impl.inc.h"
#undef PORTMUX_AQUIRE_MUX_FN_NAME
#undef PORTMUX_RELEASE_MUX_FN_NAME
#undef PORTMUX_COMPARE_SET_FN_NAME
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
esp32/esp32s2beta: Extract common SPIRAM options into esp_commmon component
2019-06-05 00:34:19 -04:00
#if defined(CONFIG_SPIRAM)
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
#define PORTMUX_AQUIRE_MUX_FN_NAME vPortCPUAcquireMutexIntsDisabledExtram
#define PORTMUX_RELEASE_MUX_FN_NAME vPortCPUReleaseMutexIntsDisabledExtram
#define PORTMUX_COMPARE_SET_FN_NAME uxPortCompareSetExtram
#include "portmux_impl.inc.h"
#undef PORTMUX_AQUIRE_MUX_FN_NAME
#undef PORTMUX_RELEASE_MUX_FN_NAME
#undef PORTMUX_COMPARE_SET_FN_NAME
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
#endif
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
#ifdef CONFIG_FREERTOS_PORTMUX_DEBUG
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
#define PORTMUX_AQUIRE_MUX_FN_ARGS portMUX_TYPE *mux, int timeout_cycles, const char *fnName, int line
#define PORTMUX_RELEASE_MUX_FN_ARGS portMUX_TYPE *mux, const char *fnName, int line
#define PORTMUX_AQUIRE_MUX_FN_CALL_ARGS(x) x, timeout_cycles, fnName, line
#define PORTMUX_RELEASE_MUX_FN_CALL_ARGS(x) x, fnName, line
#else
#define PORTMUX_AQUIRE_MUX_FN_ARGS portMUX_TYPE *mux, int timeout_cycles
#define PORTMUX_RELEASE_MUX_FN_ARGS portMUX_TYPE *mux
#define PORTMUX_AQUIRE_MUX_FN_CALL_ARGS(x) x, timeout_cycles
#define PORTMUX_RELEASE_MUX_FN_CALL_ARGS(x) x
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
#endif
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
static inline bool __attribute__((always_inline)) vPortCPUAcquireMutexIntsDisabled(PORTMUX_AQUIRE_MUX_FN_ARGS) {
esp32/esp32s2beta: Extract common SPIRAM options into esp_commmon component
2019-06-05 00:34:19 -04:00
#if defined(CONFIG_SPIRAM)
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
if (esp_ptr_external_ram(mux)) {
return vPortCPUAcquireMutexIntsDisabledExtram(PORTMUX_AQUIRE_MUX_FN_CALL_ARGS(mux));
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
}
#endif
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
return vPortCPUAcquireMutexIntsDisabledInternal(PORTMUX_AQUIRE_MUX_FN_CALL_ARGS(mux));
}
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
static inline void vPortCPUReleaseMutexIntsDisabled(PORTMUX_RELEASE_MUX_FN_ARGS) {
esp32/esp32s2beta: Extract common SPIRAM options into esp_commmon component
2019-06-05 00:34:19 -04:00
#if defined(CONFIG_SPIRAM)
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
if (esp_ptr_external_ram(mux)) {
vPortCPUReleaseMutexIntsDisabledExtram(PORTMUX_RELEASE_MUX_FN_CALL_ARGS(mux));
return;
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
}
#endif
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
vPortCPUReleaseMutexIntsDisabledInternal(PORTMUX_RELEASE_MUX_FN_CALL_ARGS(mux));
freertos: Inline vPortCPUAcquireMutex/vPortCPUReleaseMutex into implementations Further improves performance: No contention -> 134 cycles Recursion -> 117 cycles Contention -> 323 cycles
2017-02-16 00:56:06 -05:00
}
Especially when internal memory fills up, some FreeRTOS structures (queues etc) get allocated in psram. These structures also contain a spinlock, which needs an atomic-compare-swap operation to work. The psram hardware, however, does not support this operation. As a workaround, this patch detects these spinlocks and will, instead of S32C1I, use equivalent C-code to simulate the behaviour, with an (internal) mux for atomicity.
2017-12-19 02:47:00 -05:00
Reference in New Issue Copy Permalink