Read the whole memory costs additional time. When the number memory nodes is
small, test on qemu shows that read memory directly is faster.
Signed-off-by: xuxingliang <xuxingliang@xiaomi.com>
1. Avoid to_bytes by using memoryview directly.
2. No need to call gdb to cast to char *.
3. Cache the memory data without invoke gdb in every iteration.
4. Do code cleanup.
memleak speed improved from 261.93 seconds to 29.9 seconds for x4b
usecase.
Signed-off-by: xuxingliang <xuxingliang@xiaomi.com>
Now the HAVE_PFINET(6)_SOCKETS depends on NET_TCP/UDP_HAVE_STACK, which
is previously defined in net/ folder and cannot be included.
Considering many places use this check, maybe moving them to netconfig.h
could be better.
Signed-off-by: Zhe Weng <wengzhe@xiaomi.com>
Read base frequency from system counter0 and write it
to arm core register. This corrects timers to work properly
Then enable counting.
Signed-off-by: Jouni Ukkonen <jouni.ukkonen@unikie.com>
If the pointer is not set to zero, it will erroneously point to itself, resulting in an erroneous loop pointing
Signed-off-by: wangchen <wangchen41@xiaomi.com>
the stub-function entry address is stored in r4, we should branch to the
stub-function with blx r4, not r5
Signed-off-by: guoshichao <guoshichao@xiaomi.com>
/usr/bin/ld: nuttx.rel: relocation R_X86_64_32S against `.rodata' can not be used when making a PIE object; recompile with -fPIE
/usr/bin/ld: failed to set dynamic section sizes: bad value
Signed-off-by: cuiziwei <cuiziwei@xiaomi.com>
This commit adds register definitions for advanced, basic, and general purpose timers. Formatting convention was taken from the H7 hardware timer header.
These files were copied from a config of pico-sdk 2.0.0.
They provide struct address mapped access to peripherals, along
with register definitions. It also provides some compiler helpers
that are partially used by the port.
assert: tcp conn and didn't recv ack, tcp_free will assert in L909
and no need to tcp free because tcp close will free by lpwork.
Signed-off-by: meijian <meijian@xiaomi.com>
Problem Description:
Problem occurrence when Thread1 creat connect/recv socket and Thread2 cancel Thread1.
1. Thread2 cancel when Thread1 connect event, will cause DEBUGASSERT in devif_callback_free.Because cb in g_cbfreelist.
2. Thread2 cancel when Thread1 recvfrom sem-wait, when the FIN packet input and will trigger tcp_recvhandler and will crash.Becuase some thread stack data was freed.
Signed-off-by: meijian <meijian@xiaomi.com>
Hardware initialization is based refcount, not
spi enable bit and add interface to unitialize bus
Signed-off-by: Jouni Ukkonen <jouni.ukkonen@unikie.com>
reason:
Since assert may synchronously wait to stop another CPU, potentially
leading to a deadlock, we replace enter_critical_section with a
small spinlock to avoid such a situation.
Signed-off-by: hujun5 <hujun5@xiaomi.com>
Add function description for function prototype of `riscv_jump_to_user`
to make it easier to read, and fix some inconsistent comment style in
`riscv_internal.h`.
Signed-off-by: Huang Qi <huangqi3@xiaomi.com>
This reverts commit befe29801f.
Because a few regressions have been reported and
it likely will take some time to fix them:
* for some configurations, semaphore can be used on the special
memory region, where atomic access is not available.
cf. https://github.com/apache/nuttx/pull/14625
* include/nuttx/lib/stdatomic.h is not compatible with
the C11 semantics, which the change in question relies on.
cf. https://github.com/apache/nuttx/pull/14755
Both RX and TX interrupts can be enabled after the setup of the transfer; the TX interrupt
must be enabled after queuing the first byte as done before. It is not possible to miss the
RX interrupts, as it pends as long as the byte gets read from the FIFO
When starting the TX, the first byte can be queued instantly, it won't be sent out to the
bus if there is NACK to the address. This also prevents spurious TX interrupts in error
cases, since the TX queue is not empty after initiating a transfer.
In some error cases controller sends STOP by itself even if AUTOSTOP is disabled.
It is better to tell the controller to ABORT, which will also generate STOP only when needed.
Signed-off-by: Jukka Laitinen <jukkax@ssrc.tii.ae>
