Summary:
- It is applicable to esp32 products and uses the himem part
of 8M psram by creating character devices.
Impact:
- None
Testing:
- Use esp32-wrover series products for more than 1000 functional verifications.
In order to turn longjmp context-switch safe, it's necessary
to disable interrupts before modifying windowbase and windowstart.
Otherwise, after a context switch, windowstart and windowbase
would 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.
*Background:*
This PR is related to an issue first observed on ESP-IDF
https://github.com/espressif/esp-idf/issues/5229 and it was, then,
checked on NuttX using a test application.
*The test application:*
To check if the problem affects ESP32, ESP32-S2 and ESP32-S3 on
NuttX, it was created an application based on:
https://en.cppreference.com/w/c/program/longjmp
The application creates 16 tasks (`#define NUMBER_OF_TASKS 16`)
that implements the following daemon:
```
static int setjmp_longjmp_daemon(int argc, char *argv[])
{
for (int i = 0; i < NUMBER_OF_TASKS * 2; i++)
{
jmp_buf env;
volatile int count = 0;
if (setjmp(env) != UINT16_MAX)
{
foo(&env, ++count);
}
}
sem_post(&g_sem);
return EXIT_SUCCESS;
}
```
The main function also initializes a semaphore to avoid application
exiting before tasks return successfully:
```
sem_init(&g_sem, 0, -NUMBER_OF_TASKS);
```
Finally, the round-robin interval was lowered to 1ms to raise the
chances of the longjmp being interrupted by a context switch
(`CONFIG_RR_INTERVAL=1).
This setup was able to reproduce the problem prior to this patch
being applied.
heap.
QEMU had a different ROM image that used the regions of PRO CPU for both
CPUs. This was causing crashes when running SMP mode as the heap was
being corrupted when the APP CPU starts.
QEMU is now loading the same image as the hardware chip and thus this
special case doesn't exist anymore.
Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
The name used in Tensilica support file core-isa.h for all vendors is
`XCHAL_NUM_INTLEVELS`.
Use a new name may be confused by newer porting xtensa arch.
Change-Id: Ie108d3fdfcc02c81f0eacfca852a1cfc9eea17de
The PRO CPU and APP CPU have different peripherals for GPIO interrupts.
Each CPU needs to allocate an interrupt and attach it to its GPIO
peripheral.
Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
is enabled.
That region is technically part of the PRO CPU and we should be able to
allocate it early. However, QEMU uses a slightly different bootloader
image that uses the same part for both CPU. So, when APP CPU starts
during the SMP bring up it will corrupt some data.
Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
Internal heap was occupying the region straight after .data up to
HEAP_REGION1. The issue with this is if static allocation is large,
we'll end up with too little memory left for the internal heap.
Moving it to the beginning of region 2 gives us more room to play with.
Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
I skipped the following files because they were not simple.
I'll create separate PRs.
arch/xtensa/src/esp32/esp32_cpustart.c
arch/xtensa/src/common/xtensa_abi.h
boards/xtensa/esp32/esp32-core/include/board.h
Also, I skipped the following files and directories because
they looked too huge and/or foreign.
arch/xtensa/include/esp32/tie.h
arch/xtensa/include/xtensa/xtensa_corebits.h
arch/xtensa/src/esp32/hardware/
arch/xtensa/include/esp32/tie-asm.h
arch/xtensa/include/esp32/core-isa.h
arch/xtensa/include/xtensa/core.h
I also fixed a few "is is" style typos when unwrapping long lines.
