pthread_exit will be called recursive when pthread_cancel
or other cleanup operation with syscalls that support
cancellation, to avoid this by mark current tcb flag as
TCB_FLAG_CANCEL_DOING instead of TCB_FLAG_CANCEL_PENDING.
Signed-off-by: Huang Qi <huangqi3@xiaomi.com>
Drop to user-space in kernel/protected build with up_pthread_exit,
now all pthread_cleanup functions executed in user mode.
* A new syscall SYS_pthread_exit added
* A new tcb flag TCB_FLAG_CANCEL_DOING added
* up_pthread_exit implemented for riscv/arm arch
Signed-off-by: Huang Qi <huangqi3@xiaomi.com>
Previously the value of the FSInfo section was taken as granted.
As described in the white paper of FAT [1] this value is unreliable
and has to be computed at mount time.
[1] https://www.win.tue.nl/~aeb/linux/fs/fat/fatgen103.pdf
If the subcore configuration, which is mainly used in the Spresense
Arduino environment, is enabled, the serial console has been already
initialized by maincore. Then, don't need to re-initialize the UART1
serial driver.
ADC driver does not support multiple open and close. It causes the memory
corruption by multiple free. This commit fixes this problem by introducing
the reference counter.
If the system clock is changed during loading gnssfw, gnss open may be
failed. So this commit prohibits clock change until loading gnssfw is
completed.
There is an issue that the next alarm is expired immediately after
canceling a RTC alarm. Fixed alarm settings to be completely cleared
when canceling an RTC alarm.
If the system clock is changed during the SPI transfer, the SPI data can
be corrupted. So this commit prohibits the clock change during SPI transfer,
and keep the clock until the transfer is completed.
Introduce PM_CPUFREQLOCK_FLAG_HOLD into the frequency lock mechanism in
power manager, which is used to keep the current frequency without clock
change, for example, during the transfer of a periphral.
UART driver is stopped and re-started during a clock change. When a UART
interrupt is generated in each process, the unexpected behavior will
occur and a console will get stuck with UART driver. This commit fixed
each process is performed atomically.
Broken by 635cfa.
On an stm32 the heap is formed from
const uintptr_t g_idle_topstack = HEAP_BASE;
where HEAP_BASE is &_bss + CONFIG_IDLE_STACKSIZE.
Both these values are not deterministic. One
comes from the compiler, the other the system
configurator.
a 3 byte bss and 250 byte stack would lead to
and unaligned address used as the heap to be.
The compiler used clever `strd r1,r3,[r5,#8]` to
store 2 values in one memory cycle into the
heap_impl struct. Resulting in a hardfault.
Change the amount of bss or the CONFIG_IDLE_STACKSIZE
could lead to a non-functional NuttX system.
Fix bug that select() did not return when the nfds argument
was set to a negative value. The specification is that -1 is
set to the return value and EINVAL is set to errno.
