and removed two tcp_send_txnotify() calls from tcp_sendfile (they are not needed anymore).
As a result, the TX throughput of both the tcp_send_buffered and tcp_send_unbuffered
is significantly boosted in case of TUN/TAP network device.
For Linux, we rename the section so that they are not invoked.
However, it doesn't work for macOS.
Notes:
* Renaming the sections (as the Linux version does) doesn't work
because the section is looked by its flags, not by its names.
* No convenient tools available to alter the relevant bit in
the section flags.
* The constructors are invoked before the program entry point.
In case of high TX network traffic, netdriver_loop() that reads data from netdev
was invoked via up_idle() only after high TX network traffic had stopped.
That resulted in massive delay and drop of TCP ACK packets and
any other packets from netdev (tun/tap device).
For example, task is blocked by nxsem_wait(sem1), use nxsem_wait(sem2)
in signal handler, and take sem2 successfully, after exit from signal
handler to task, nxsem_wait(sem1) returns OK, but the correct result
should be -EINTR.
Signed-off-by: Zeng Zhaoxiu <zhaoxiu.zeng@gmail.com>
we have many different hostfs implementation now, so it's better
to select the implementation explicitly, just like what we have
done for arm(FS_HOSTFS vs. ARM_SEMIHOSTING_HOSTFS).
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Gregory Nutt has submitted the SGA
Omni Hoverboards has submitted the SGA
David Sidrane has submitted the ICLA
Mateusz Szafoni has submitted the ICLA
Sebastien Lorquet has submitted the ICLA
Paul Alexander Patience has submitted the ICLA
as a result we can migrate the licenses to Apache.
Signed-off-by: Alin Jerpelea <alin.jerpelea@sony.com>
This reverts commit 45672c269d.
Because:
* It's very confusing to have cc as LD.
* I don't see what "-nostartfiles -nodefaultlibs" in LDFLAGS are
supposed to do when we use LD directly. It would be simpler to
remove them from our LDFLAGS.
When SIM_SPI is valid, a specified Linux SPI device ‘spidevN.P’(N is bus number and P is CS number) is attached to nuttx simulator, shown as 'spi0' under /dev. One may type spi command (need SPITOOL valid) in NSH to control the Linux SPI and exchange data, other devices such sensors can use it to debug in simulator on a Ubuntu PC. Note that a USB<>SPI module (e.g. CH341A/B) should be plugged in to achieve Linux SPI ports.
Change-Id: I275b2c2bbf6d14bcdf514c89efb9a2264d69e9a3
Signed-off-by: liucheng5 <liucheng5@xiaomi.com>
When use poweroff command,host_abort will be called,but may be make __stack_chk_fail irq, and host_abort be called in PANIC(), so bring infinite loop, in turn it can not exit SIM.
Signed-off-by: buyuer <dingddding@163.com>
so the source code compiled by host environment can include config.h
directly and then avoid pass Kconfig option through Makefile manually
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: Ic9fe6f846082cef2d0808dc717df8ae6ed929edf
Summary:
- I noticed that sim:smp does not work correctly on macOS
due to the recent changes
- Actually, it can not receive the IPI host signal, so if
a new task is scheduled on CPU1/2/3, it hangs.
- Finally, I found that sim_cpu_start() calls pthread_attr_init()
which locates not on the host OS but on the NuttX
- This commit fixes this issue by adding the symbol names
to nuttx-names.in
Impact:
- sim in SMP mode
- NOTE: stack usage for CPU1/2/3 IDLE is still incorrect on macOS
Testing:
- Tested with ostest on Ubuntu 18.04 (x86_64) and macOS 11.4 (x86_64)
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
since this macro isn't check from host side after:
commit cee43ce280
Author: Xiang Xiao <xiaoxiang@xiaomi.com>
Date: Fri Jul 16 02:23:59 2021 +0800
arch/sim: Initialize the idle thread stack info correctly
and change the default value of IDLETHREAD_STACKSIZE to 65536
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Summary:
- I noticed that sim:smp does not work correctly on macOS
due to the recent changes
- Actually, it can not receive the IPI host signal, so if
a new task is scheduled on CPU1/2/3, it hangs.
- Finally, I found the issue depends on pthread stack settings
and perhaps it might affect the host signal handling.
- This commit fixes this issue by just reverting the pthread
stack setting only for macOS.
Impact:
- sim:smp on macOS
- Stack usage for CPU1/2/3 IDLE will be incorrect
Testing:
- Tested with ostest on macOS 11.4 (x86_64)
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
Cleaning during `clean_context` had the issue of remaking everything
when `menuconfig` was issued. That's because `menuconfig` has a
`clean_context` on its way.
Signed-off-by: Abdelatif Guettouche <abdelatif.guettouche@espressif.com>
use the same condition check in declaration and reference
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: I7b05316e914708fceeddac394d784ee3720a3c1b
and change the default value of IDLETHREAD_STACKSIZE to 65536
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: Ia54efbbca4b69706150bc4178844b316688a104e
1.Move all special process to mm_takesemaphore
2.Remove the support of recurive lock
3.Remove mm_trysemaphore function
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: Ie216a6294ab67c5d427f31b089beb15c532f08fe
it's more simple to make mm_heap_s opaque outside of mm
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: I5c8e435f6baba6d22b10c5f7e8d9191104fb5af2
and remove mm_sbrk and kmm_sbrk since it's wrong to expose
sbrk to other heaps except the default userspace heap.
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Root casue:
when do thread exit, need add free stack operation to mm_delaylist,
but in SMP mode, CPU0 thread1 exit, at this time, CPU1 call malloc
and free mm_delaylist.
Fix:
Divide mm_delaylist for per CPU in SMP mode.
Change-Id: Ibf7d04614ea2f99fb5b506356b7346a0d94f0590
Signed-off-by: ligd <liguiding1@xiaomi.com>
since kernel component should use UTC instead local time
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: Icf939e1ab0af8e577105f539d2553bc67b3b3d10
instead calling kmm_heapmember or umm_heapmember because:
1.The stack supplied by caller may allocate from heap too
2.It's hard to implement these two function in ASan case
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
Change-Id: I196377822b7c4643ab4f29b7c1dc41dcd7c4dab1
All modern desgin support stack pointer and it's also an
important information, so let's standardize this interface.
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
