The current essid setting(SIOCSIWESSID) are not satisfactory in case
of specified bssid connection, if there are multiple essids with same name,
we need additional bssid information to connect to the specified softap.
The extended flag "WAPI_ESSID_DELAY_ON" instructs the driver to delay the
connection behavior of essid, so that which can accept more accurate
information before generating a connection.
About flow of wapi command, drivers that support WAPI_ESSID_DELAY_ON semantics
will have the following behavior changes:
1. Station mode without bssid set:
$ ifup wlan0
$ wapi mode wlan0 2
$ wapi psk wlan0 12345678 3
$ wapi essid wlan0 archer 1
$ renew wlan0
2. Station mode with bssid set:
$ ifup wlan0
$ wapi mode wlan0 2
$ wapi psk wlan0 12345678 3
$ wapi essid wlan0 archer 2 <----- WAPI_ESSID_DELAY_ON will indicate the driver delay
$ the connection event late to bssid set (SIOCSIWAP).
$ wapi ap wlan0 ec:41:18:e0:76:7e
$ renew wlan0
Signed-off-by: chao.an <anchao@xiaomi.com>
wget is missing from some system (like macOS and Windows native),
it's better to use curl to simplify build environment.
Signed-off-by: Huang Qi <huangqi3@xiaomi.com>
Therefore, the associate can support the connection of specified
channel if the scan results contain the multiple AP with same essids.
Change-Id: I4e657c52e638693e7ae19aea2a4fe8a431885547
Signed-off-by: chao.an <anchao@xiaomi.com>
Summary:
- This commit enables gs2200m application to get IWNWID/IWFREQ/IWSENS information
Impact:
- Affects gs2200m only
Testing:
- Tested with spresense:wifi
Summary:
- This commit adds support for ioctl(fd, SIOCGIFADDR, ...) to gs2200m_main.c
Impact:
- Only affects ioctl(fd, SIOCGIFADDR, ...) with gs2200m
- Need to update nuttx as well
Testing:
- Tested with spresense:wifi
- Tested with dhcpc
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
Summary:
- This commit adds getperrname_request() to gs2200m_main.c
Impact:
- Affects gs2200m only
- Need to update nuttx as well
Testing:
- Tested with spresense:wifi
- Modify telnetd and add getpeername() to show a client address
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
Summary:
- This commit fixes to handle address info in accept()
Impact:
- All use cases which use accept() with gs2200m
- Need to update nuttx as well
Testing:
- Tested with spresene:wifi
- Tested with telnet daemon
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
Summary:
- This commit fixes to handle UDP connect() with bind() to local port
- NOTE: GS2200M does not support TCP connect() with bind to local port
Impact:
- All UDP cases which use connect() with gs2200m
- Need to update nuttx as well
Testing:
- Tested with spresense:wifi
- Create a UDP socket and bind() to local port.
- Then connect() to remote address with port and send()
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
Summary:
- During Wi-Fi stress test, I noticed ASSERT happens in gs2200m_main.c
- Actually, req->max_buflen was 0 but the case was not handled correctly
- This commit fixes this issue
Impact:
- Affects read/recvfrom with len=0 cases
Testing:
- Tested with spresense:wifi_smp
- Run nxplayer for audio streaming
- Run ps/free/ifconfig commands via telnet
- Run smp command via telnet
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
This commit changes only ones with the default 2048 and
leaves the others.
E.g. this leaves SYSTEM_RAMTEST_STACKSIZE, whose default is 1024.
I guess those need to be inspected one-by-one.
For now, I left the following instances because it isn't
clear to me why they are using the different values.
Maybe they need one-by-one inspection.
examples/igmp/Makefile:STACKSIZE = 1024
examples/powerled/Makefile:STACKSIZE = 1024
examples/powermonitor/Makefile:STACKSIZE = 768
examples/relays/Makefile:STACKSIZE = 512
examples/smps/Makefile:STACKSIZE = 1024
graphics/screenshot/Makefile:STACKSIZE = 4096
system/flash_eraseall/Makefile:STACKSIZE = 1024
testing/cxxtest/Makefile:STACKSIZE = 4096
testing/smart_test/Makefile:STACKSIZE = 4096
The ASSERT is too strong, as it cannot be disabled, so changing to
DEBUGASSERT, as calling these functions should not be a problem as
long as the application logic handle the error correctly.
The error code is changed to ENOSYS, as it seems to better reflect
the fact that the call itself would be valid, but the functionality
is not implemented (see [1] or `man errno`).
[1] https://www.gnu.org/software/libc/manual/html_node/Error-Codes.html
Author: Alan Carvalho de Assis <engenharia03@siam.ind.br>
Run nxstyle again gs2200m_main.c file and fix issues
Author: YAMAMOTO Takashi <yamamoto@midokura.com>
gs2200m: Retry indefinately on GS2200M_IOC_ASSOC failure
The ioctl doesn't return why it failed.
It might or might not be a transient failure.
In my environment, gs2200m often returns the following for AT+WA.
It usually works after a few retries.
[ 12.110000] _parse_pkt_in_s1: +++++ 0:(msize=19, msg=WLAN CONNECT ERROR|)
[ 12.110000] _parse_pkt_in_s1: +++++ 1:(msize=6, msg=ERROR|)
Run all .c and .h files in last PR through nxstyle and fix all complaints.
Author: Xiang Xiao <xiaoxiang@xiaomi.com>
Wapi improvement (#16)
* wireless: add passphrase control
* wireless: add disconnect control
1.Remove void cast for function because many place ignore the returned value witout cast
2.Replace void cast for variable with UNUSED macro
Change-Id: Ie644129a563244a6397036789c4c3ea83c4e9b09
Signed-off-by: Xiang Xiao <xiaoxiang@xiaomi.com>
1. Check NSH_NETINIT for self network initialization
2. Check NSH_ARCHINIT for sel arch specific initialization
3. Always show help regardless of NSH_BUILTIN_APPS
4. Loop forever regardless of NSH_BUILTIN_APPS, user could:
a.change the default behavior by the command line argument
b.or ctrl+c to break out the loop
i8shark: Adds support for intentionally suppressing passing the FCS so that Wireshark doesn't try to validate it.
Approved-by: Gregory Nutt <gnutt@nuttx.org>
I8sak improvements
* wireless/ieee802154: Adds support for getting/setting maxretries and promsicuous mode from i8sak
* wireless/ieee802154/i8sak: Fixes a build warning and handles a few cases where a fault can occur due to not catching missing arguments.
Approved-by: Gregory Nutt <gnutt@nuttx.org>
