esp_execute_shared_stack_function always restored the stack watchpoint
regardless of CONFIG_FREERTOS_WATCHPOINT_END_OF_STACK. This would lead
to an abondoned but active watchpoint on a former stack once the task
calling esp_execute_shared_stack_function is deleted, if
CONFIG_FREERTOS_WATCHPOINT_END_OF_STACK is inactive.
This has been fixed now.
Closes https://github.com/espressif/esp-idf/issues/10414
- Add support for esp32s2, esp32c3 and esp32c2 for the `memprot`-related tests
- Preliminary support for esp32s3 has also been added,
the test app will be enabled for esp32s3 later when
the memprot-related issues are fixed.
- Override panic handler to dump the violation intr status
- Dump the `memprot` violation registers before calling the
real panic handler
- Handle `Illegal Instruction` exception in case of memprot permission violation
* In esp32c3 with `memprot` enabled, if we try to execute arbitrary code
from RTC_FAST_MEM we get an `Illegal Instruction` exception from the panic
handler rather than a `Memory Protection Fault`.
* This is because the Illegal Instruction interrupt occurs earlier than the
memory protection interrupt due to a higher interrupt latency.
- Added minor improvements to `panic` test app
* Replaced existing API to disable flash cache which did not disabled cache always
(`esp_flash_default_chip->os_func->start(esp_flash_default_chip->os_func_data)`)
with `spi_flash_enable_interrupts_caches_and_other_cpu`
* Included some required headers explicitly (`esp_memory_utils.h` and `esp_heap_caps.h`)
All the partition handling API functions and data-types were moved from the 'spi_flash' component to the new one named 'esp_partition'. See Storage 5.x migration guide for more details
pytest_panic used to do 'from conftest import PanicTestDut'. This
stopped working when another conftest.py file was added in
tools/test_build_system/conftest.py.
In this case we can't rename conftest.py since both are necessary
for pytest to install the hooks in the respective test cases.
Fix by moving PanicTestDut into a separate module, then importing
it from there.
This commit fixes build issue because of function `bootloader_common_reset_rtc_retain_mem`
getting inlined with compiler optimization level set to `PERF` (-O2).
Build failure log:
-----------------
In function 'bootloader_common_reset_rtc_retain_mem',
inlined from 'bootloader_common_update_rtc_retain_mem' at /h/esp-idf/components/bootloader_support/src/bootloader_common_loader.c:183:13:
/h/esp-idf/components/bootloader_support/src/bootloader_common_loader.c:159:5: error: 'memset' writing 16 bytes into a region of size 0 overflows the destination [-Werror=stringop-overflow=]
159 | memset(rtc_retain_mem, 0, sizeof(rtc_retain_mem_t));
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
This commit marks all functions in soc_hal.h and soc_ll.h as deprecated.
Users should use functions from esp_cpu.h instead.
Also added missing wrap funcions for esp_cpu_stall() in test_panic.c files.
In C++20, using the result of an assignment to a 'volatile' value is
deprecated.
Breaking change: register "setter" or modification macros can no
longer be used as expressions.
Closes https://github.com/espressif/esp-idf/issues/9170
This updates the minimal supported version of CMake to 3.16, which in turn enables us to use more CMake features and have a cleaner build system.
This is the version that provides most new features and also the one we use in our latest docker image for CI.
After refactoring the target components (e.g. esp32) no longer contained any real functionality.
What remained in these components have been moved elsewhere and the component itself deleted from the
build system.
When vfs component is not added to the build, bare minimum syscalls
are provided by newlib component. These syscalls currently don't
perform CR/LF translation.
This commit makes the 'get_test_name' function work with minimal
syscalls and also adds echo, so that the user sees what they type.
lwip was added to common requirements list to provide "sys/socket.h"
header to all components without additional requirements specified.
However, lwip pulls in a lot of dependencies on other components.
This commit removes lwip from common requirements to reduce the number
of components in G1-only apps.
To compensate for this removal, the following changes are made:
- newlib (which is a common requirement) has a public dependency on
lwip if lwip is present in the build. This ensures that sys/socket.h
is available as long as lwip component is included into the build.
- lwip is now a public requirement of esp-tls since esp_tls.h includes
sys/socket.h header.
- lwip is now a public requirement o esp_http_client because
sys/socket.h is included from esp_http_client.h
- lwip is now a private requirement of esp_wifi for "smartconfig_ack"
- lwip is now a private requirement of mqtt for socket functions
- lwip is now a public requirement of tcp_transport because
esp_transport_tcp.h includes sys/socket.h header.
- mbedtls checks if lwip component is present in the build. If yes,
net_sockets.c is added to the build, along with the dependency on
lwip. Previously lwip was a public requirement of mbedtls
unconditionally.
system/g1_components test app is updated to reflect the changes
Default public dependencies of a component before and after this
change, except common requirements:
- esp_timer (public dependency of freertos)
- bootloader_support (public dependency of esp_hw_support)
- vfs (public dependency of lwip)
- esp_wifi (public dependency of lwip)
- esp_event (public dependency of esp_wifi)
- esp_netif (public dependency of esp_event)
- esp_eth (public dependency of esp_netif)
- esp_phy (public dependency of esp_wifi)
After:
- esp_timer (public dependency of freertos)
- bootloader_support (public dependency of esp_hw_support)
Altogether, the following components have been always added as
public requirements to all other components, and are not added now
([breaking-change]):
- lwip
- vfs
- esp_wifi
- esp_event
- esp_netif
- esp_eth
- esp_phy
Application components now need to explicitly declare dependencies on
these components.
Some components were including esp_timer.h without declaring a
dependency on esp_timer component. This used to work due to a
transitive public dependency on esp_timer from freertos component.
Add explicit dependencies where needed.
Also some source files were using esp_timer functions without
including the header file. This used to work because esp_timer.h was
included from freertos port header file. This commit adds esp_timer.h
includes where needed.
This commit updates the visibility of various header files and cleans up
some unnecessary inclusions. Also, this commit removes certain header
include paths which were maintained for backward compatibility.
For some of the MSPI high frequency setting (e.g. 80M DDR mode Flash or PSRAM), timing tuning is required.
Certain delays will be added to the MSPI RX direction. When system clock switches down, the delays should be
cleared. When system clock switches up, the delays should be restored.
Replaced hyphens with underscores in examples
project definition for all examples which had
hyphens in their project name. dpp-enrollee is
an exceptions because the name matches the
project directory name while the project
directory also contains hyphens.
* On S3, the placement of ROM functions is
ECO-dependent. Hence, we don't jump into
the middle of the longjmp function in ROM
on S3 anymore.
Instead, the whole longjump function is used
in the patch.
* Also properly excluded the patch from
bootloader build with Makefiles
Closes IDF-3391
The previous location was the return from the first ets_printf call
that prints ROM sign-on message. Since the main function was patched
in ECO3, the new address no longer works — there is no instruction at
0x40007901 in ECO3 ROM. This could be solved by setting two
breakpoints (one would work for ECO <=2, the other for ECO3), but we
would need to remove the unused breakpoint later.
Fix this by setting the breakpoint at ets_printf. This means that when
debugging a loadable ELF the ROM sign-on message will no longer be
shown, but this doesn't seem to be an issue.
* Patched longjmp to be context-switch safe
longjmp modifies the windowbase and windowstart
registers, which isn't safe if a context switch
occurs during the modification. After a context
switch, windowstart and windowbase will be
different, leading to a wrongly set windowstart
bit due to longjmp writing it based on the
windowbase before the context switch. This
corrupts the registers at the next window
overflow reaching that wrongly set bit.
The solution is to disable interrupts during
this code. It is only 6 instructions long,
the impact shouldn't be significant.
The fix is implemented as a wrapper which
replaces the original first instructions of
longjmp which are buggy. Then, it jumps back
to execute the rest of the original longjmp
function.
Added a comparably reliable test to the
test apps.
By unchecking "Place panic handler code in IRAM" in the menuconfig,
the panic handlers will be placed in flash. Of course, flash cache must
be activated when entering panic handlers.
Theory is that on the runner, in rare cases, gdb may need more than
1 second to load and start responding to commands.
However it's possible these timeouts are due to some other problem
(like gdb failing)
The previous approach was to allocate an array on the stack, and
have the array extend past the stack size. This worked by would
result in SP being moved near the end of the stack. If an interrupt
triggered at that time, interrupt prologue would try to save the
context to the stack, tripping the stack overflow watchpoint.
Replacing this with the approach which doesn't move the SP and simply
writes to decreasing addresses from SP, until stack overflow check
triggers.
Small quality-of-life improvement, make it easier to run specific
test cases, when debugging the tests locally.
Take the optional list of test cases to run from the command line.
