SDMMC databook specifies that when an error interrupt is generated
(such as RTO, DCRC, RCRC), CMD_DONE interrupt is also generated.
The code used to break out of the event handling loop to wait for
another interrupt event to happen, reporting CMD_DONE.
However if interrupt processing was delayed, it is possible that
a single event contains both RTO and CMD_DONE. Previously, CMD_DONE
would not be handled, and the state machine would be stuck in
SDMMC_SENDING_CMD state, until a timeout. This didn't change the
outcome (err=0x107), but delayed the handling of response timeout
event.
Fix by not breaking out of the event handler, optionally processing
the CMD_DONE interrupt if it has been reported in the same message.
...to avoid defining common symbols.
GCC since version 10 defaults to -fno-common and doesn't generate
common symbols, leading to duplicate definitions of these symbols.
GCC since version 10 uses -fno-common by default and will not emit
common symbols. Enable this option to find the occurrences of common
symbols in ESP-IDF.
Closes https://github.com/espressif/esp-idf/issues/5080
Can be used in target-specific C code to distinguish when Linux target
is used.
Also don't enable CONFIG_APP_BUILD_GENERATE_BINARIES and
CONFIG_APP_BUILD_BOOTLOADER when building for Linux, since the
bootloader is not used. CONFIG_APP_BUILD_GENERATE_BINARIES might make
sense in some cases, can be re-enabled later.
When building for "linux" (~POSIX) target on macOS, the system linker
is normally used. MacOS linker doesn't recognise --gc-sections, but
has a -dead_strip flag which is equivalent.
This fixes passing -Wimplicit-fallthrough=0 flag to Clang, which it
doesn't recognize, and is not necessary for this file.
Also the flag is changed from PUBLIC to PRIVATE since it is necessary
when compiling this component only.
The compiler was not informed that the assembly block should be
treated as a memory barrier and could optimize out the initialization
of local arrays which could be used as semihosting arguments.
This resulted in garbage values being passed as semihosting call
arguments.
Additionally this commit changes the approach for placing values into
specific register. Instead of clobbers, local register variables are
used. This results in simpler generated code, since the compiler is
able to place values directly into the registers used for semihosting
arguments, avoiding additional moves.
In C++ code it is considered to be undefined behavior to exit a
non-void function without returning a value. Normally this is
detected by the compiler, but users could disable relevant warnings.
Add a note about this possibility.
See https://github.com/espressif/esp-idf/issues/8244 for context.
To make the transition from 32-bit time_t to 64-bit time_t smoother,
detect the size of this type in CMake and remove the manual option in
Kconfig.
The information about 64-bit time_t support is moved from Kconfig help
string into the "system time" section of the API reference.
...and all their callers.
With the upcoming switch from sizeof(time_t)==4 to sizeof(time_t)==8,
sizeof(struct stat) is also increasing.
A few newlib functions present in ROM allocate 'struct stat' on the
stack and call _fstat_r on this structure. The implementation of
fstat is provided in ESP-IDF. This implementation will often do
memset(st, 0, sizeof(*st)), where st is 'struct stat*', before setting
some fields of this structure. If IDF is built with sizeof(st)
different from sizeof(st) which ROM was built with, this will lead
to an out-of-bounds write and a stack corruption.
This commit removes problematic ROM functions from the linker script.
Here are the functions which allocate 'struct stat':
* _isatty_r (in ROM)
* __swhatbuf_r, called by __smakebuf_r, called by __swsetup_r and
__srefill_r (in ROM)
* _fseeko_r (not in ROM)
* glob2 (not in ROM)
* _gettemp (not in ROM)
As a result, these functions are used from libc.a, and use correct
size of 'stat' structure.
Closes https://github.com/espressif/esp-idf/issues/7980
- Check if pip is installed for sys.executable before attempting to
create the virtual environment, bail out with an error if not.
- Don't pass --seeder argument to virtualenv if its version is
too old. For example, on Ubuntu 18.04, virtualenv 15.1.0 doesn't
support this argument.
- Pass --python argument to virtualenv to request specific interpreter
to be used.
Closes https://github.com/espressif/esp-idf/issues/8045
- add missing Wi-Fi header MD5 checks for esp32c3, esp32s3
- check PHY libraries for references to ets_printf, in addition to
checking Wi-Fi libraries. This used to happen until libphy.a was
moved into a separate submodule.
Reported in https://github.com/espressif/esp-phy-lib/issues/3
Some tests check if certain files are rebuilt when source files are
'touch'ed. With ccache, 'touch'ing source files doesn't cause a
rebuild, hence the test fails. In case IDF_CCACHE_ENABLE was set in
the environment, unset it before starting the tests.
xPortStartScheduler calls vPortSetupTimer -> _frxt_tick_timer_init,
which enables tick timer interrupt and sets up the first timeout.
From that point on, the interrupt can fire. If the interrupt happens
while _frxt_dispatch is running, the scheduler will enter an infinite
loop. This is because _frxt_dispatch isn't supposed to be preemptable,
and the tick interrupt will overwrite some of the registers used by
_frxt_dispatch.
Note that this situation doesn't practically occur on the real
hardware, where the execution of vPortSetupTimer and _frxt_dispatch
happens quickly enough. However it can be reproduced on an emulator
if the tick period is set to 1ms.
Add an explicit call to portDISABLE_INTERRUPTS in xPortStartScheduler
to guarantee that _frxt_dispatch doesn't run with interrupts enabled.
This is similar to the esprv_intc_int_set_threshold(1); call in
RISC-V version of port.c.
COMPONENT_DIRS and EXTRA_COMPONENT_DIRS should be defined as CMake
lists, using 'set' or 'list' commands. Some applications written
for earlier versions of ESP-IDF used to define these variables as
space separated strings.
For example, the following is correct:
set(EXTRA_COMPONENT_DIRS path/to/components path/to/more/components)
The following is not correct:
set(EXTRA_COMPONENT_DIRS "${EXTRA_COMPONENT_DIRS} component1")
set(EXTRA_COMPONENT_DIRS "${EXTRA_COMPONENT_DIRS} component2")
The string "component1 component2" may indicate a single directory
name with a space, or two directory names separated by space.
However due to the fact that such way of defining EXTRA_COMPONENT_DIRS
was supported in IDF 4.3 and earlier, we need to provide backward
compatibility for it.
This commit introduces a new script, split_paths_by_spaces.py, which
is invoked if EXTRA_COMPONENT_DIRS or COMPONENT_DIRS variable contains
spaces. The script tries to determine if each space should be
interpreted as a separator or as part of the directory name.
When this cannot be done unambiguously, the script reports an error.
In all cases when space separators are detected, the script reports
a warning, and prints instructions for fixing the CMakeLists.txt.
Breaking change in this commit: specifying non-existent directories
in COMPONENT_DIRS or EXTRA_COMPONENT_DIRS is no longer allowed.
Fixes WORKING_DIRECTORY argument being passed with spaces escaped `\ `
which causes failure in run_cmd.cmake, since the WORKING_DIRECTORY is
interpreted as having a literal backslash character in it.
New —-list-separator argument of confgen.py and
prepare_kconfig_files.py is used to select which character is used
as list separator. For compatibility with esp-docs, we still keep
support for space separator. Otherwise esp-docs would have to choose
the separator depending on the IDF version.
* Adds implementations of __{atomic,sync}_nand_fetch_n. These builtins
were implemented for other operations but were not defined for NAND.
* Adds implementation of __atomic_OP_fetch_n for all OPs.
* Adds implementation of __sync_OP_and_fetch_n for all OPs.
Reported in https://github.com/espressif/arduino-esp32/issues/5948
It can be used in component CMakeLists.txt files when adding "weak"
dependencies between component. A "weak" dependency from A to B is
similar to target_link_libraries(A PRIVATE B), but it gets added only
if B is included in the build.
Use it instead of specifying B as part of PRIV_REQUIRES when the
component can be built (even if with reduced functionality) without
component B being available.
Ivan Grokhotkov