Rpmsg Port Uart is a new rpmsg transport layer.
Just like the rpmsg port spi, the difference is that the physical
communication method changed from SPI to UART.
Signed-off-by: Bowen Wang <wangbowen6@xiaomi.com>
Add get_local_cpuname to the rpmsg framework ops to support communicate
with the same remote core with multi rpmsg transport.
Some rpmsg services will send local cpu name to remote core and then let
remote core to connect local core by using this cpu name, when there are
multi rpmsg channels with same remote core, the remote core may connect
to incorrect core, so use the error rpmsg channel.
For example, there are two rpmsg channels between ap and audio:
ap core audio core
[ap1] <-- rpmsg virtio1 --> [audio1]
[ap2] <-- rpmsg virtio2 --> [audio2]
When we want to use the rpmsg virtio1 to communicate, ap core may send
local cpuname "ap2" to audio, so the audio core use remote cpu "ap2" to
connect with ap, and resulting in the use of incorrect rpmsg channel.
Signed-off-by: Bowen Wang <wangbowen6@xiaomi.com>
Rpmsg Router is new rpmsg transport layer, it can router the rpmsg
messages to a cpu that not directly connected with local cpu by Rpmsg,
For the rpmsg services, it is as if there is a real Rpmsg Channel between
the local cpu and the remote cpu.
For examples, there are three cpus: ap, cp and audio.
ap and cp, ap and audio has share memory and be connected by Rpmsg VirtIO,
so ap and cp, ap and audio can communicate with each other by Rpmsg, but
cp can not communicate with audio direclty.
[cp] <-- rpmsg virtio --> [ap] <-- rpmsg virtio --> [audio]
With rpmsg router, the cp can communicate with audip by Rpmsg dereclty because
the router in ap will forward the rpmsg message from cp/audio to audio/cp, like
this:
+<----- rpmsg router --> hub <-- rpmsg router ------>+
| | |
[cp] <-- rpmsg virtio --> [ap] <-- rpmsg virtio --> [audio]
Signed-off-by: yintao <yintao@xiaomi.com>
Add Rpmsg-Port-SPI transport layer.
Rpmsg Port SPI is a new rpmsg transport layer based on the Rpmsg Port,
it provides the capability for two SPI-connected (and two extra GPIO)
chips to communicate with each other using Rpmsg.
All already implemented Rpmsg Services can be used with this new transport
layer without any modifications.
Signed-off-by: liaoao <liaoao@xiaomi.com>
Rpmsg Physical Transport Layer is a new rpmsg transport, it's a
common part for the physical communication based rpmsg transport
layers such as Rpmsg-SPI and Rpmsg-Uart.
It implements three common parts:
1. Implement the NuttX and OpenAMP rpmsg frameworks' ops and the
rpmsg name service;
2. The buffer management and provide some APIs to lower layer to use;
Signed-off-by: liaoao <liaoao@xiaomi.com>
This interface may be used by application specific logic to start
specified slave cpu core under the pseudo AMP case which is different
with armv7-a/armv8-a SMP. Support for this function is required by
board-level logic if CONFIG_BOARDCTL_START_CPU is selected.
Signed-off-by: wanggang26 <wanggang26@xiaomi.com>
add support for UART 16550 compatible PCI device.
For now we support qemu serial PCI devices and AX99100 based cards.
Signed-off-by: p-szafonimateusz <p-szafonimateusz@xiaomi.com>
prepare 16550 UART driver to support PCI:
- [breaking change] change argument of uart_ioctl() from `struct file *filep` to `FAR struct u16550_s *priv`
Also fix moxart_16550.c build related to this change
- [breaking change] change argument of uart_getreg() and uart_putreg from `uart_addrwidth_t base` to `FAR struct u16550_s *priv`
Also fix arch/x86/src/qemu/qemu_serial.c and arch/x86_64/src/intel64/intel64_serial.c related to this change
- [breaking change] change argument of uart_dmachan() from `uart_addrwidth_t base` to `FAR struct u16550_s *priv`
- move `struct u16550_s` to public header
- generalize UART_XXX_OFFSET so we can use it with any register increment
- make u16550_bind(), u16550_interrupt(), u16550_interrupt() public
- remove arch/or1k/src/common/or1k_uart.c and use common 16550 MIMO interfacve
- change irq type in `struct u16550_s` from uint8_t to int to match MSI API
Signed-off-by: p-szafonimateusz <p-szafonimateusz@xiaomi.com>
1. only support pci modern device;
2. need the pci controller support MSI/MSI-X;
It has been verified based on virtio-rng and virtio-net.
Signed-off-by: Yongrong Wang <wangyongrong@xiaomi.com>
Signed-off-by: Bowen Wang <wangbowen6@xiaomi.com>
rpmsg_virtio_ivshmem is a ivshmem based rpmsg virtio driver,
with this driver, we can use the rpmsg virtio in qemu platfrom
Signed-off-by: Yongrong Wang <wangyongrong@xiaomi.com>
Signed-off-by: Bowen Wang <wangbowen6@xiaomi.com>
Rpmsg VirtIO is a virtio transport implementation for Rpmsg, and
it's different to the rptun framework.
rpmsg_virtio.c implements the rpmsg virtio transport layer by itself
to avoid use the remoteproc implementation in OpenAMP to save code
size, so it can be treated as a lightweight version of rptun.
Therefore, rpmsg_virtio.c only support the communication feature and
do not support contoll the life cycle of the remote core.
But benefit by it's small footprint, it can be used in the chips with
small flash.
Signed-off-by: Yongrong Wang <wangyongrong@xiaomi.com>
Signed-off-by: Bowen Wang <wangbowen6@xiaomi.com>
Some devices need to clear the fb panbuf when waking up from sleep
to avoid outputting residual buffers from before sleeping
Signed-off-by: rongyichang <rongyichang@xiaomi.com>
IOB buffer participates in the calculation of the congestion strategy
in the protocol stack, if the iob buffer is exhausted or there are
too many syslog printing tasks, the behavior of the protocol stack
and log printing will unable to determined, this PR will remove support
for iob buffer in syslog
Signed-off-by: chao an <anchao@lixiang.com>
According to the mmap(2) specification, anonymous pages should be initialized to zero unless the MAP_UNINITIALIZED is specified.
Signed-off-by: ouyangxiangzhen <ouyangxiangzhen@xiaomi.com>
This commit add MADV_HUGEPAGE and MADV_NOHUGEPAGE definitions for madvise. Notice that none of these flags have been implemented yet.
Signed-off-by: ouyangxiangzhen <ouyangxiangzhen@xiaomi.com>
coresight_claim_device will fail when unregister device
which has been enabled. device should be the init state when
it is unregistered.
Signed-off-by: liaoao <liaoao@xiaomi.com>
64-by-32-bit divisions are prominent in the NuttX, even on 32-bit
machines. Luckily, many of them use a constant divisor that allows
for a much faster multiplication by the divisor's reciprocal.
The compiler already performs this optimization when compiling a 32-by-32
division with a constant divisor. Unfortunately, on 32-bit machines, gcc
does not optimize 64-by-32 divisions in that case, except for constant
divisors that happen to be a power of 2.
Let's avoid the slow path whenever the divisor is constant by manually
computing the reciprocal ourselves and performing the multiplication
inline. In most cases, this improves performance of 64-by-32 divisions
by about two orders of magnitude compared to the __div64_32() fallback,
especially on architectures lacking a native div instruction.
Signed-off-by: ligd <liguiding1@xiaomi.com>
Bowen Wang
liaoao
yintao
p-szafonimateusz
wanggang26
dongjiuzhu1
Xiang Xiao
zhanghu5
wangjianyu3
wangyongrong
ligd
fangpeina
ouyangxiangzhen
rongyichang
chao an
zhangyuan29