apache
netdev_upperhalf: convert Kconfig to variable
Summary
convert Kconfig to variable, upperhalf can supports multiple working modes at the same time, which is specified by NIC when register, the multi-network card scenario will be more flexible.
Impact
All network card drivers implemented using struct netdev_lowerhalf_s.
It is necessary to initialize appropriate rxtype and priority according to the business scenario during network card initialization.
Testing
sim:matter with ping and iperf2.
nsh> ifconfig
eth0 Link encap:Ethernet HWaddr 42:00:d7:82:13:6b at RUNNING mtu 1500
inet addr:10.0.1.2 DRaddr:10.0.0.1 Mask:255.255.255.0
inet6 addr: 3380::4000:d7ff:fe82:136b/64
inet6 DRaddr: fe80::a096:6ff:fe65:382a
lo Link encap:Local Loopback at RUNNING mtu 1518
inet addr:127.0.0.1 DRaddr:127.0.0.1 Mask:255.0.0.0
inet6 addr: ::1/128
inet6 DRaddr: ::1
nsh> ping -c 5 10.0.1.1
PING 10.0.1.1 56 bytes of data
56 bytes from 10.0.1.1: icmp_seq=0 time=0.0 ms
56 bytes from 10.0.1.1: icmp_seq=1 time=0.0 ms
56 bytes from 10.0.1.1: icmp_seq=2 time=0.0 ms
56 bytes from 10.0.1.1: icmp_seq=3 time=0.0 ms
56 bytes from 10.0.1.1: icmp_seq=4 time=0.0 ms
5 packets transmitted, 5 received, 0% packet loss, time 5050 ms
rtt min/avg/max/mdev = 0.000/0.000/0.000/0.000 ms
nsh> ipe
iperf
iperf2
nsh> iperf2 -c 10.0.1.1 -i 1 -t 5
------------------------------------------------------------
Client connecting to 10.0.1.1, TCP port 5001
TCP window size: 16.0 KByte
------------------------------------------------------------
[ 2] local 10.0.1.2 port 17206 connected with 10.0.1.1 port 5001
[ ID] Interval Transfer Bandwidth
[ 2] 0.00-1.00 sec 49.1 MBytes 412 Mbits/sec
[ 2] 1.00-2.00 sec 49.2 MBytes 412 Mbits/sec
[ 2] 2.00-3.00 sec 48.8 MBytes 409 Mbits/sec
[ 2] 3.00-4.00 sec 49.8 MBytes 418 Mbits/sec
[ 2] 4.00-5.00 sec 51.6 MBytes 433 Mbits/sec
[ 2] 0.00-5.01 sec 249 MBytes 416 Mbits/sec
nsh>
nsh> iperf2 -s -i 1
------------------------------------------------------------
Server listening on TCP port 5001
TCP window size: 16.0 KByte
------------------------------------------------------------
tcpnodelay failed: Unknown error 38
[ 3] local 10.0.1.2 port 5001 connected with 10.0.1.1 port 60610
[ ID] Interval Transfer Bandwidth
[ 3] 0.00-1.00 sec 44.9 MBytes 377 Mbits/sec
[ 3] 1.00-2.00 sec 43.9 MBytes 368 Mbits/sec
[ 3] 2.00-3.00 sec 49.6 MBytes 416 Mbits/sec
[ 3] 3.00-4.00 sec 47.6 MBytes 399 Mbits/sec
[ 3] 4.00-5.00 sec 48.6 MBytes 407 Mbits/sec
[ 3] 0.00-5.00 sec 235 MBytes 394 Mbits/sec
I'm confused about what's been removed? Is there no longer a netdev work thread? Why? Am I correct in thinking that users can't choose between HPWORK and LPWORK queues anymore?
Why not split the commits into multiple PRs? The PR title is misleading since this is much more than modifying a Kconfig variable. Can you please update it?
@linguini1 As mentioned in the Summary, it is change the configuration from defconfig to be specified by the code when the network card is initialized. The relevant configuration methods have also been updated in the documentation.
dev->rxtype = NETDEV_RX_WORK;
dev->priority = HPWORK;
Now rx direct has been removed from patch chain. Previously, I just wanted to synchronously demonstrate how to do the related extension support in the new upperhalf.
The relevant configuration methods have also been updated in the documentation.
I don't see any documentation updates in this PR. Is it in a separate PR?
dev->rxtype = NETDEV_RX_WORK; dev->priority = HPWORK;
This looks fine but is it accessible from the external API when registering? Sorry I haven't looked very closely at the code change yet.
The relevant configuration methods have also been updated in the documentation.
I don't see any documentation updates in this PR. Is it in a separate PR?
dev->rxtype = NETDEV_RX_WORK; dev->priority = HPWORK;This looks fine but is it accessible from the external API when registering? Sorry I haven't looked very closely at the code change yet.
He included the documentation at Documentation/components/net/netdriver.rst but as you noticed it is missing the new API: netdev_upper_exit_thread()
The relevant configuration methods have also been updated in the documentation.
I don't see any documentation updates in this PR. Is it in a separate PR?
dev->rxtype = NETDEV_RX_WORK; dev->priority = HPWORK;This looks fine but is it accessible from the external API when registering? Sorry I haven't looked very closely at the code change yet.
He included the documentation at Documentation/components/net/netdriver.rst but as you noticed it is missing the new API: netdev_upper_exit_thread()
@acassis but netdev_upper_exit_thread is an internal static function, which isn't callable from lowerhalf driver.
The relevant configuration methods have also been updated in the documentation.
I don't see any documentation updates in this PR. Is it in a separate PR?
dev->rxtype = NETDEV_RX_WORK; dev->priority = HPWORK;This looks fine but is it accessible from the external API when registering? Sorry I haven't looked very closely at the code change yet.
He included the documentation at Documentation/components/net/netdriver.rst but as you noticed it is missing the new API: netdev_upper_exit_thread()
@acassis but netdev_upper_exit_thread is an internal static function, which isn't callable from lowerhalf driver.
Hmm, ok in this case it is not a visible API. Thank you @xiaoxiang781216