Startup flow refactoring
This set of commits changes the startup code in a way that lets the application choose if/when to initialize WiFi/BT.
Application entry point is now a more familiar `main()` function. This function is executed in its own task. Application may choose to do some initialization from main function, create some tasks and then return from `main`. Simple applications may choose to do all their work from `main`.
Additionally this MR splits event handling code into two parts.
- One part is a set of standard handlers for WiFi and DHCP events. Most applications will use this set of handlers, and it is made available via new `esp_event_process_default` function.
- Another part is the default implementation of event handling loop. Some applications may choose to use default event loop through `esp_event_loop_` set of APIs, which start an event handling task and call user-provided event callback from this task. Other applications may create an event queue and implement event loop themselves. In this case application has to provide `esp_event_send` function. In this case the implementation provided by `esp_event_loop_` module is unused.
esp-idf-template has been updated to match this set of changes: https://github.com/espressif/esp-idf-template/tree/feature/init_refactoring
BT example has also been updated.
We need to provide examples of both event handling approaches. This will be done in a separate follow-up MR.
See merge request !112
Fix a spelling mistake in panic message, add carriage return to stack canary message
Probably the most trivial mr ever, but the spelling error was reported as a bug and the missing \r bugged me ever since I saw it.
See merge request !113
* master:
components/spi_flash: remove stray level of indentation
components/nvs: fix broken sentences in comment blocks
Roll back submodule version
Spinlocks already come initialized. Remove the code that would essentially re-initialize them at runtime
Remove all references to prvLockQueue / prvUnlockQueue
components/esp32: clean up unused function warnings in single core mode
clean up warnings
components/nvs: fix build, use log library instead of printf
components/spi_flash: add flash operation counters option to Kconfig
components/nvs: add erase function
components/nvs: fix formatting
components/nvs: batch writes when possible
components/spi_flash: add performance counters
components/nvs: maintain item hash list at page level
components/nvs: avoid reading just-erased page
# Conflicts:
# components/esp32/cpu_start.c
# components/esp32/event_default_handlers.c
nvs: speed improvement and erase APIs
NVS initialization speed issue raised in https://ezredmine.espressif.com/issues/6881 has been addressed for single core mode through optimization of `spi_flash_*` routines. This MR addresses same issue for dual core mode.
- An index table is added at Page level, it allows doing fast search of item index based on key name and namespace. This heavily reduces the number of flash reads.
- Write operations are batched when writing/erasing long variable-length values (str and blob). This gives a 33% reduction in the number of writes for a 64-byte value, 50% reduction for a 96-byte value.
- Add optional (configurable via menuconfig) perfomance counters for spi_flash APIs
With these changes, total init time for NVS goes down from 2 seconds to 140 ms.
Additional RAM usage is 128-640 bytes per page, depending on content. Typical RAM usage for current use case (storing wifi configuration parameters) is 256 bytes.
This change also exposes APIs to erase single key or the whole namespaces, needed for https://ezredmine.espressif.com/issues/6769.
See merge request !105
Clean up warnings and remove unused parts of FreeRTOS
For third party components (lwip and expat), compilation flags are adjusted to silence existing warnings (i have manually checked that all warnings are benign).
In components/esp32, replaced use of WIFI_DEBUG with ESP_LOG functions.
Additionally prvLockQueue/prvUnlockQueue have been removed, dragging some other bits of FreeRTOS along.
Original issue: https://github.com/espressif/esp-idf/issues/18
See merge request !111
- put contents of a few headers into c++ guard blocks
- fix off-by-one error in do_global_ctors
- remove system_init from startup code (should be called from main)
This change separates definitions in esp_event.h and functions in event.c into several parts:
- event structure definitions (esp_event.h)
- default implementations of event handlers (event_default_handlers.c)
- default implementation of event loop (event_loop.c, esp_event_loop.h)
Purpose of this change is to allow applications choose their own poison:
- full control of event loop at the expense of more bootstrap code
- pre-defined event task firing event callbacks, but less code in app_main.c
For third party components (lwip and expat), compilation flags are adjusted to silence existing warnings (i have manually checked that all warnings are benign).
In components/esp32, replaced use of WIFI_DEBUG with ESP_LOG functions.
Only remaining warning is in FreeRTOS queue.c, and it may be a useful one.
This change removes implicit WiFi/BT initialization from startup code.
"main" task is started once essential part of startup code is complete.
This task calls application-provided "int main(void)" function, which can call WiFi/BT init functions if necessary.
1. add SSL session new and free function
2. add SSL session peer cert get and free operation
3. above all, change low-level cert object to be object point not object
For third party components (lwip and expat), compilation flags are adjusted to silence existing warnings (i have manually checked that all warnings are benign).
In components/esp32, replaced use of WIFI_DEBUG with ESP_LOG functions.
Only remaining warning is in FreeRTOS queue.c, and it may be a useful one.
Introduces new internal function, Page::alterEntryRangeState, which gathers changes to multiple elements of entry state table into a single write, provided that these changes fall into a single word. This allows changing state of up to 16 entries in a single write.
Also adds new function, writeEntryData, which writes the whole payload of SZ and BLOB type entries in one go, instead of splitting it into multiple 32-byte writes.
This reduces number of writes required for SZ and BLOB entries.
1. add empty function to load private key into SSL context
2. add empty function to load certification into SSL context
3. add function to load RSA private key
1. add openssl 'new' and 'free' function
2. add clear unused variate to void warning to appear when compile
3. add internal function 'X509_new' to take the place of 'sk_X509_NAME_new_null' function whitch is openssl stack function
This feature allows to use static buffers (or from a pool of memory which is not
controlled by FreeRTOS).
In order to reduce the impact of the changes, the static feature has only been added
to the queus (and in consequence to the semaphores and the mutexes) and the tasks.
The Timer task is always dynamically allocated and also the idle task(s), which in the
case of the ESP-IDF is ok, since we always need to have dynamic allocation enabled.
lwip: add debug code to show udp/tcp pcbs
Add code to show all tcp/udp pcbs, these kind of debug info is helpful for lwip issue debugging.
See merge request !98
* master: (57 commits)
components/lwip: fix grammar
components/lwip: make SO_REUSE configurable via menuconfig
bootloader: remove trailing newlines from log messages
components/freertos: override per-task __cleanup handler to close stdin, stdout, stderr
components/esp32: move peripheral symbols to separate ld script
components/log: regroup macros for better readability
gitlab-ci: allow running tests for branches, triggered via API
components/log: fix timestamp calculation
components/log: set default runtime log level to ESP_LOG_VERBOSE
components/log: fix error when using ESP_LOGx from C++ code
components/log: fix bugs, add options to override log level for files, components, and bootloader
fix ledc and spi typo
remove prefix and postfix
Enable SO_REUSEADDR in LWIP
freertos: fix memory debug macro issue Define configENABLE_MEMORY_DEBUG according to CONFIG_ENABLE_MEMORY_DEBUG
peripheral structure headers: move volatile keyword from members to typedef
Adding -fstrict-volatile-bitfields to the CFLAGS/CXXFLAGS. Without this, gcc tries to access bitfields using the smallest possible methods (eg l8i to grab an 8-bit field from a 32-bit). Our hardware does not like that. This flag tells gcc that if a bitfield is volatile, it should always use the type the field is defined at (uint32_t in our case) to size its access to the field. This fixes accessing the hardware through the xxx_struct.h headers.
add peripheral module struct headers
build system docs: Add note about no spaces in component names
Docs: Add note about unusual submodule messages when cloning on Windows
...
# Conflicts:
# components/esp32/cpu_start.c
# components/esp32/include/soc/cpu.h
Add logging library
Logging library, intended to be used across other components.
Uses design similar to Android NDK logging APIs and allows for both compile time and run time filters, similar to logcat.
Also includes cleanup of cpu_startup.c — i was changing some logging output in this file so decided to re-format it and reduce code duplication.
Ref TW6703
See merge request !68
* branch 'master':
components/freertos: override per-task __cleanup handler to close stdin, stdout, stderr
gitlab-ci: allow running tests for branches, triggered via API
Enable SO_REUSEADDR in LWIP
Daniel initially asked me why this wasn't enabled, and I don't think I got any reasons against enabling this. If any, it makes porting existing software easier. Tuan needs it for UDP multicast as well.
Code changes are by Tuan: basically the enable for SO_REUSEADDR in LWIP as well as a bugfix in a bit of mbedTLS that gets enabled.
See merge request !90
Move CPU region protection setup into soc/cpu.h
change tabs to spaces
remove unused extern declarations
use RTC_WDTCONFIG0 instead of numeric address (still need to fix BB reg)
This also removes logging implementation from bootloader and replaces it
with the one provided by the log component. Some occurrences of printf
and ets_printf have been changed to ESP_LOGx APIs.
Note that with WiFi stack enabled, system_init will reset frequency to 240MHz.
To make this setting useful, esp32-wifi-libs submodule needs to be updated.
esp32/lib: update to 1303c92c
1. support to set vendor ie;
2. fix "mode:" log twice issue; TW7019
3. fix station can not connect to some AP in WEP mode; TW7021
See merge request !85
Function declaration should only be included if MBEDTLS_SHA512_ALT is
not set. This matches sha1.h and sha256.h
This change should be contributed back upstream to mbedTLS project.
* branch master: (51 commits)
fix typos
callback: add a void* pointer with each callback
header: callbacks use "void *" as arguments
header: format and fix typos
add wifi api comments
add wifi api comments
components/spi_flash: call SPIUnlock only once
components/spi_flash: improve comments and readability
esp32 cpu_start: Include rom/uart.h, remove inline ROM function declarations
bootloader: Add bootloader Kconfig
Fix bug where nesting vTaskEnterCritical calls would not re-enable interrupts after vTaskExitCritical sections.
installed gitlab key in global before_script, don't need to overwrite in template before_script
ci: Use github repo for wifi libraries on master branch, gitlab repo otherwise
components/esp32,spi_flash: update some comments
components/esp32: remove dependency of WIFI_ENABLED on FREERTOS_UNICORE
esp32/lib: Bump WiFi revision
update ci: build ssc with idf submodule wifi lib
components/esp32: add inter-processor call API and implement spi_flash through it
add MACRO to get flash pad number from efuse value
add comments to register read/write operations
...
lwip: add socket tx flow control
This fix is used to solve wifi q full issue:
1. When wifi tx q is full, set tx_enable to false
2. When wifi tx q decrease 1/2 max length, set tx_enable to true again
3. esp_wifi_tx_enable return whether tx_enable is true
4. If esp_wifi_tx_enable return false, socket will be blocked
See merge request !62
This fixes the performance impact for spi_flash_write and spi_flash_erase.
With this change, NVS init in single core mode takes about 50ms (compared to >2seconds before that).
Dual core compatible spi flash APIs
Implementation of spi flash APIs which work in dual core mode.
This also includes other changes:
- fixes to FreeRTOS tasks.c made by Jeroen and me — these were originally made to fix issues with high priority tasks seen in WiFi stack, but were required to get stuff in ipc.c working as well
- initial base idea with two high-priority tasks was separated into a separate API — esp_ipc.h. That API allows one to run a function on the other CPU from a high-priority task. Because we don't have preemption across CPUs, this is currently not very fast: it requires on average half a tick until the second CPU switches to the high priority task. Currently that API can be either blocking, or non-blocking with queue length of 1. I haven't yet found a use case for arbitrary sized queue of IPC requests. Anyone who needs that can trivially add that later by replacing a binary semaphore with a generic queue.
With these changes i can run application with WiFi stack and NVS on both cores.
See merge request !25
1. timer reg file for both time group 0 and time group 1, not only timer group 0
2. fix bug that io mux header file mismatch with chip
3. fix bug that some BASE address not correct
4. add some static function to eagle.fpga32.rom.addr.v7.ld
5. add interrupts usage table
6. add some comments for rom code functions
* master: (130 commits)
lwip: Define LWIP_ESP8266 in port lwipopts.h not gcc command line
CI: Build the esp-idf-template with the matching branch name, if it exists
README: Add Resources section with some links
Rename README.buildenv to docs/build_system.rst and ReST-ify it
Eclipse docs: Easier to just replace entire PATH, msys32 has everything we need to build/flash
test_build_system: Print ESP_IDF_TEMPLATE_GIT for easier debugging
Name component makefiles component.mk instead of Makefile
Eclipse doc: Add troubleshooting note about Makefile directories
eclipse_make.sh: Fix printing of make directory
Move bin/eclipse_windows_make.sh to tools/windows_eclipse_make.sh
Eclipse docs: Prepend IDF paths to beginning of PATH
Set default SPI flash access mode to DIO
FreeRTOS: temporary solution for memory canaries and memory debug
tcpip_adapter: fix dhcp client work flow
event: not post got ip event if static is invalid
tcpip_adapter: typedef clean up
event: post got ip event when use static ip
tcpip_adapter: use dhcp callback to post got ip event
dhcp: add dhcp callback
lwip: remove netif_reg_addr_change_cb
...
Used by mbedTLS to set MBEDTLS_CONFIG_FILE in all components.
This change sets CFLAGS/etc at the project level and then exports those
variables for components, rather than setting them independently each time
a component Makefile is invoked.
Eclipse fixes, including renaming component Makefiles to component.mk
Some changes after following up some Eclipse problems with @nilay .
Biggest thing is renaming all component Makefiles to component.mk. This requires changes to all esp-idf projects, ie:
* https://github.com/espressif/esp-idf-template/compare/bugfix/docs_eclipse?expand=1
* http://gitlab.espressif.cn/idf/esp-idf-tests/tree/update/component_makefiles
... this fixes Eclipse weirdness and also means the only "Makefile"s in the project are the ones you should run make against.
If this MR is merged then I'll also merge these, email all developers about the breaking change and post on esp32.com.
See merge request !34
Set default SPI flash access mode to DIO
We have a bunch of boards out there (the NodeMCU-lookalike, breadboardable boards) in the hands of early reviewers. Because the flash chips on these boards have a different command to go into QIO mode, ESP-IDF will break on these boards out of the box. Eventually, we need some nifty way to autodetect this or something, but for now, the quick fix is to make DIO the default access mode.
Once this is merged, I will also modify the default in the template apps sdkconfig.
See merge request !65
Should protect against concurrent use of hardware crypto primitives,
with good performance.
Not necessary to call esp_aes_acquire_hardware(),
esp_sha_acquire_hardware(), etc when using these APIs. These are
provided for external users calling the hardware crypto hardware
directly, to coexist with this implementation.
Fixes problems with Eclipse trying to build in directories it shouldn't.
This is a breaking change for existing repositories, they need to rename
any component Makefiles to component.mk and rename their references to
$(IDF_PATH)/make/component.mk to $(IDF_PATH)/make/component_common.mk
FreeRTOS: temporary solution for memory canaries and memory debug
1. This is just a temporary solution, it will be removed when umm_malloc is ready
2. Support memory canaries mechanism
2. Add debug code to show allocated memory info
See merge request !64
tcpip_adapter: got ip event stuff
Include:
1. add dhcp callback;
2. post got ip event in dhcp callback when dhcp bind;
3. tcpip_adapter typedef clean;
4. post got ip event in station connected event handler when use valid static ip;
TW6926
See merge request !61
1. This is just a temporary solution, it will be removed when umm_malloc is ready
2. Support memory canaries mechanism
2. Add debug code to show allocated memory info
Since esp32 has much more memory than esp31B, modify the HEAP_HIGHWAT from
6K to 20K.
Notes: HEAP_HIGHWAT is used for tcp/udp tx flow control, if socket task detects
that the free heap memory is less than HEAP_HIGHWAT, it stops to trasmitting until
the free heap memory is greater than HEAP_HIGHWAT
1. Add the parameter type of api esp_wifi_set_storage.
2. WIFI_STORAGE_RAM means storing profiles in ram. WIFI_STORAGE_FLASH means storing profiles in flash.
Fix: Return after setting breakpoint when ocd is detected instead of continuing …
…into the panic routine
Made a small thinko when refactoring the panic handler code. This fixes it.
See merge request !36
The parameter aid followed -a is aid of the connected station. If aid is 0, kick out all of the connected stations.
2. Add block parameter to api esp_wifi_scan_start.
If block is true, the api blocks before scan done or returns directly.
components/expat: add XML parse feature
Develop and Compile XML Parse based on XMPP protocol, can test from unity running list, see expat file.
See merge request !32
found small mistake in it
we have:
PROVIDE ( ets_aes_crypt = 0x4005c9b8 );
PROVIDE ( ets_aes_disable = 0x4005c8f8 );
PROVIDE ( ets_aes_enable = 0x4005c8cc );
PROVIDE ( ets_aes_set_endian = 0x4005c928 );
PROVIDE ( ets_aes_setkey_dec = 0x4005c994 );
PROVIDE ( ets_aes_setkey_enc = 0x4005c97c );
Signed-off-by: Ivan Grokhotkov <ivan@espressif.com>
Implement lwip per-thread-sem with:
vTaskSetThreadLocalStoragePointerAndDelCallback/pvTaskGetThreadLocalStoragePointer
register sys_thread_tls_free to free per-thread-sem when the thread is terminated.
Also let lwip use index CONFIG_LWIP_THREAD_LOCAL_STORAGE_INDEX
2. Add struct station_list into tcpip_adapter layer
3. When ap -L cmd is received, get mac address from mac layer first and then search ip address based on mac address on dhcp layer.
update ld and heap region
Divide origin fpga ld to several lds, choose specific ld for normal/BT/Trace by menuconfig, change iram/dram len in ld to full use hardware ram region.
See merge request !28
Fix dualcore startup
This MR includes:
- a workaround for a hardware bug with cache initialization
- fix of `#if CONFIG_WIFI_ENABLED` block in cpu_start.c for the case when WiFi is disabled.
- removal of miscellaneous delays in cpu_start.c
See merge request !22
fix DRAM_ATTR attribute
Variables marked with DRAM_ATTR were not placed in any specific section and were appended to the last section, causing some errors at binary file generation stage.
This change places contents of .dram1 section into .dram0.data.
Also removes .iram1pro/.iram1app/.dram1pro/.dram1app which weren't used anywhere.
See merge request !23
Make: fix incremental builds, add build tests
Four semi-related build changes in one:
* Add basic tests for make system sanity (can be expanded as we find corner cases)
* Fix incremental building of bootloader when source files change
* Don't regenerate component libraries or re-link the ELF file if nothing changed
* Rename `$(vecho)` to `$(summary)` and add a new `$(details)` macro to echo some helpful build details when V=1.
See merge request !20
This is a workaround for a hardware bug with cache initialization.
Only two lines of code related to DPORT_APP_CACHE_MMU_IA_CLR were added around mmu_init(1); call,
and two lines at the end of comment block.
I reformatted surrounding lines to use spaces for indentation, like the rest of the code in this file does.
This change removes various delays in PRO and APP CPU startup paths.
Also some unused code and ets_printf statements are cleaned up or removed.
2nd CPU now communicates that it has started through a global variable.
The thread-local-storage feature in FreeRTOS attaches an application-usable array of pointers to a thread control block. These pointers usually point to a structure the thread allocates. When a thread gets (voluntarily or involuntarily) destroyed, this memory can leak. This merge adds a matching second array of user-settable pointers to destructor routines. As soon as the task gets cleaned up (which happens in the idle thread), the destructors get called and the memory can be freed.
See merge request !19
Add cr after a lf is printed (configurable)
Small menuconfig option to make the newlib stdout functions replicate the behaviour of ets_printf e.a. by sending out a \r before every \n.
See merge request !14
This commit fixes several issues with state handling in nvs::Page. It also adds extra consistency checks in nvs::PageManger initialization.
These changes were verified with a new long-running test ("test recovery from sudden poweroff"). This test works by repeatedly performing same pseudorandom sequence of calls to nvs_ APIs. Each time it repeats the sequence, it introduces a failure into one of flash operations (write or erase). So if one iteration of this test needs, say, 25000 flash operations, then this test will run 25000 iterations, each time introducing the failure point at different location.
rtos: change XT_CLOCK_FREQ from 13MHz to 80MHz
Freq has been change to `80MHz` in system_init, otherwise system tick will not be accurate now.
Maybe we can config freq by menuconfig.
See merge request !8
