This commit adds a workaround for the TWAI listen only mode errata which is
present on the ESP32, ESP32-S2, ESP32-S3, and ESP32-C3. twai_get_status_info()
has also been updated to account for the fact that TEC/REC are frozen in
listen only mode.
Errata Description:
When the TWAI controller is put into listen only mode, it should not influence
the TWAI bus in any way (i.e., should never send a dominant bit). However,
on the targets listed above, the TWAI controller will send dominant bits in an
error frame (i.e., active error frame), even if the controller is set to listen
only mode.
Workaround:
We can force the TWAI controller into the error passive state on startup (by
setting the REC to >= 128). Since the TEC/REC are frozen in listen only mode,
the TWAI controller will remain error passive and only send recessive bits
(i.e., passive error frames), thus will not influence the TWAI bus.
Closes https://github.com/espressif/esp-idf/issues/9157
When light (or comprehensive) poisoning is enabled, the size requested by the user for allocation
is extended by a few bytes to store the canary header and footer. heap_caps_get_allocated_size() should
return the original size asked by the user (without the additional canary bytes).
test_malloc.c extended with a new test assuring that heap_caps_get_allocated_size() returns the proper size
regardless of the degree of poisoning.
When stack check is enabled, certain functions (sometimes placed in RAM)
are being decorated with stack guards and a call to __stask_chk_fail() in
case ofr stack corruption. For this reason, __stack_chk_fail() must be
placed in RAM too.
Add stack check config in heap tests on all targets to find eventual flash to RAM
calls due to stack checks when running callgraph_check.py
On xtensa architecture, the call to __assert_func uses a reference to __func__ that can
sometimes be placed in flash. Since the __asert_func can be called from functions in IRAM
the check_callgraph script can report an error when checking for invalid calls from IRAM
to flash sections. However, the __asert_func prevents this scenario at runtime so the
check_callgraph script reports a 'flas positive' situation. For this reasson, all references
to __func__$x found prior to a call to __assert_func are droped in the parsing of the rtl files.
this commits:
- adds build-time test to check that no call to flash regions are done from IRAM functions
- resolves problems related to IRAM function using content in flash memory
- update heap_caps_alloc_failed to use a default function name in DRAM
when necessary instead of creating a function name variable in DRAM for
each call of heap_caps_alloc_failed. This allows to save some extra bytes
in RAM.
This commits adds a internal.md file in the heap directory to clarify the idea behind
which functions is placed in IRAM or in flash.
A section in mem_alloc.rst documentation is added to specify which functions from the
heap component API can be used in interrupt handlers.
This commit aims to place in the IRAM section only the functions that
are relevent for performance instead of placing the entire content of
multi_heap.c, mullti_heap_poisoning.c and tlsf.c in the IRAM.
The calculation of fl index max is changed to always be the smallest
number that includes the size of the registered memory.
The control_construct() function now checks for minimum size as the control structure
parameters are calculated.
There is no longer a minimum configuration for fl index max so the tlsf_config
enum is striped down to remove unecessary compile time values.
the tlsf_size() function will fail if no tlsf pointer is passed as parameter since there
is no way to calculate a default tlsf size anymore.
bitfields are now used in control_t when possible which reduces the size of the structure
from 56 bytes to 36 bytes.
When SECURE BOOT V2 is enabled and CONFIG_SECURE_BOOT_SIGNED_BINARIES
is not set, sector-pad the bootloader image, which is required for an
external PKCS#11 interface to generate a signature.
esptool_py: Update submodule to release/v3 (4bc311767b7c6df41def6f95a50f87b1c9406cbd)
Before this fix, when we call esp_pm_configure after gpio_wakeup_enable,
the configuration of GPIO in sleep state in gpio_wakeup_enable will be
overwritten by esp_pm_configure.
mktime function uses tm_isdst member as an indicator whether the time
stamp is expected to be in daylight saving time (1) or not (0).
FAT filesystem uses local time as mtime, so no information about DST
is available from the filesystem.
According to mktime documentation, tm_isdst can be set to -1, in which
case the C library will try to determine if DST was or wasn't in
effect at that time, and will set UTC time accordingly.
Note that the conversion from UTC to local time and then back to UTC
(time_t -> localtime_r -> FAT timestamp -> mktime -> time_t) does not
always recover the same UTC time. In particular, the local time in the
hour before DST comes into effect can be interpreted as "before DST"
or "after DST", which would correspond to different UTC values. In
this case which option the C library chooses is undefined.
Closes https://github.com/espressif/esp-idf/issues/9039
Originally reported in https://github.com/espressif/arduino-esp32/issues/6786
1. fix ftm timer macro issue
2. send null data when state change from association to run
3. bugfix for optimize abnormal beacon log
4. bugfix RSNXE related changes
5. send ftm frames immediately
6. fix ftm procedure with peer failed status 4 issue
For Secure Boot v2 case, unsigned image is first padded to next 64K
aligned boundary and then a signature block of 4K gets appended. Thus
an app partition whose size is 4K aligned should be allowed here.
For Secure Boot v1 case, app partition size must be 64K aligned as the
signature block lies at the very end of 64K boundary.
Relevant:
57b601ab7f
- If internet connectivity weakened or disappeared suddenly while we were
in the TLS handshake stage, the app got stuck at that point indefinitely.
- This was because when timeout was not explicitly specified in esp_tls_cfg_t,
the default timeout was set at the wrong place. This causes the sockets to be
setup with zero timeout, hence the indefinite wait.
- Add config to set duplicate scan list refresh period
- overwrite the oldest device infor if the list is full
- Fixed duplicate scan period is not accurate
- Add config to set duplicate scan list refresh period
- overwrite the oldest device infor if the list is full
- Fixed duplicate scan period is not accurate
- Fixed some memory was not released after bluetooth controller initialization failed
- Fixed privacy errorr for directed advertising
- Fixed multiple periodic advertising sync failed in air
- Fixed lld_con.c line 3048 assert
- Fixed crash sometimes when connected as a slave by the 8th device
1 change default TX power to 9 dBm
2 fixed disconnect reason 0x1f (unspecified error)
3 fixed connection timeout due to terminate ind has not been acknowledged
1.wifi: fixed sniffer mode can not capture control packet when station is enabled
2.Fixed WAPI issue on GTK compability and reference to broadcast arp pkts.
3.Fixed TX fragment issue that PN code is less than MSDU after PN code add two for TX fragment.
4.Added station connect retry count to make it easier to connect to the except AP
5.Fixed some issues on Soft-AP power save.
6.Fixed SoftAP memory leak after wifi init and deinit
7.Fixed sta get ap info does not return connected ssid issue.
8.Fixed SoftAP disconnection happened even when receive PS-Poll.
Bluetooth low power related logic and regs have separate power domain from MAC and BB,
and do not power down during light sleep. If reset when power up MAC and BB in sleep
flow, it may destroy the state of bt low power part.
By default, disabled the BLE Encrpyption requirement for provisioning characteristic.
With this flag enabled, when remote attempts to read and if the ACL link is not encrypted,
ESP device will return Insufficient Authentication. It is remote device responsibility to go
for link encryption which may result in pairing.
Some devices do not proceed for any pairing and just show failure pop-up. Also, user needs
to remove bonding on remote phone manually and then try again. This is causing bad user experience.
End user can enable it as per their use case.
- Added config to disable scan backoff
- Fixed llm_scan.c assert at line 1485 during controller deinit if duplicate scan is not stopped
- Call pll track in controller task
In previous transmission(s), the TX_DONE interrupt raw bit may be raised, but never been cleared.
TX_DONE interrrupt status bit should be cleared before enabling it to check the new transmission.
Since cpu retention dma use rc fast as clk source, so rc_fast_digi
will be enabled when we config to pd cpu. And cpu retention does not
need rc fast keep on during light sleep. So, if we use rc_fast_digi
to determine whether rc fast can be powered down, then cpu and and
rc fast cannot pd at the same time.
introduced in e44ead5356
1. The int8M power domain config by default is PD. While LEDC is using
RTC8M as clock source, this power domain will be kept on.
But when 8MD256 is used as RTC clock source, the power domain should
also be kept on.
On ESP32, there was protection for it, but broken by commit
e44ead5356. Currently the power domain
will be forced on when LEDC is using RTC8M as clock source &&
!int8m_pd_en (user enable ESP_PDP_DOMAIN_RTC8M in lightsleep). Otherwise
the power domain will be powered off, regardless of RTC clock source.
In other words, int8M domain will be forced off (even when 8MD256
used as RTC clock source) if LEDC not using RTC8M as clock source, user
doesn't enable ESP_PDP_DOMAIN_RTC8M, or in deep sleep.
On later chips, there's no such protection, so 8MD256 could't be used as
RTC clock source in sleep modes.
This commit adds protection of 8MD256 clock to other chips. Fixes the
incorrect protection logic overriding on ESP32. Now the power domain
will be determiend by the logic below (order by priority):
1. When RTC clock source uses 8MD256, power up
2. When LEDC uses RTC8M clock source, power up
3. In deepsleep, power down
4. Otherwise determined by user config of ESP_PDP_DOMAIN_RTC8M,
power down by default. (This is preferred to have highest
priority, but it's kept as is because of current code structure.)
2. Before, after the macro `RTC_SLEEP_CONFIG_DEFAULT` decides dbias, the
protection above may force the int8m PU. This may cause the inconsistent
of dbias and the int8m PU status.
This commit lifts the logic of pd int8m/xtal fpu logic to upper layer
(sleep_modes.c).
Related: https://github.com/espressif/esp-idf/issues/8007, https://github.com/espressif/esp-idf/pull/8089
temp
bluetooth: use non-blocking coexistence callback functions so that Wi-Fi task is not blocked by Bluetooth(backport v4.3)
See merge request espressif/esp-idf!20780
light sleep wakeup flag is true to indicate the most recent successful wakeup from light sleep,
which means the most recent light sleep occurred successfully and then wakes up by wakeup source
