sergiotarxz/nuttx
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
33,018 Commits 1 Branch 0 Tags 291 MiB
4cb61bd8c2
Commit Graph

1573 Commits

Author SHA1 Message Date
Gregory Nutt
09d5d05b95 net/TCP: Extend the TCP notification logic logic so that it will also report loss of connection events. 2018-09-09 17:32:10 -06:00
Gregory Nutt
9d3148406c Signals were not a good choice of IPC to implement the poll function for several reasons: In order to handle the asynchrnous poll-related event, a substantial amount of state information is needed. Signals are only capable of passing minimal amounts of data. There are also complexities with performing kernel space signal handlers in kernel space code that is better to avoid. So, instead of signals, the equivalent logic was converted to run via a callback that executes on the high-priority work queue. 
Squashed commit of the following:

    Fix up some final compile isses.

    net/netdev:  Convert the network down notification logic to use the new wqueue-based notification factility.

    net/udp:  Convert the UDP readahead notification logic to use the new wqueue-based notification factility.

    net/tcp:  Convert the TCP readahead notification logic to use the new wqueue-based notification factility.

    mm/iob:  Convert the IOB notification logic to use the new wqueue-based notification factility.

    sched/wqueue:  Signals are not good IPCs to support the target poll functionality for several reasons including the amount of data that can be passed with a signal and in the fact that in protected and kernel modes, user threads executing signal handlers in protected, kernel memory is problematic.  Instead, convert the same logic to perform the notifications via function callback on the high priority work queue.
 2018-09-09 15:01:44 -06:00
Gregory Nutt
20814acad2 sched/signal: In signal notification facility, use sigqueue() to notify vs. kill(). With sigqueue, we can pass more info (but still not enough). 2018-09-09 11:57:25 -06:00
Gregory Nutt
32e3e51678 net/netdev: Add signal notification for the case where the network goes down. 2018-09-09 10:39:25 -06:00
Gregory Nutt
28f73bd928 net/tcp and udp: Add logic to signal events when TCP or UDP read-ahead data is buffered. 
Squashed commit of the following:

    net/tcp:  Add signal notification for the case when UDP read-ahead data is buffered.  This is basically of clone of the TCP notification logic with naming adapted for UDP.

    net/tcp:  Add signal notification for the case when TCP read-ahead data is buffered.
 2018-09-09 09:21:39 -06:00
Gregory Nutt
bb92b98dc0 net/arp/arp_out.c: Fix some backward logic in an if condition. 2018-08-29 10:51:44 -06:00
dongjianli
09c689ea15 net/icmp and icmpv6: Fix failure to poll ICMP socket issue. 2018-08-26 15:37:03 -06:00
Xiang Xiao
40b0e98d25 net/icmp/icmp_conn.c and net/icmpv6/icmpv6_conn.c: Change the type of id in icmpv[6]_findconn to uint16_t 2018-08-26 15:28:02 -06:00
dongjianli
23f14c2660 net/usrsock: Add the listen/accept/getpeername/ioctl support 2018-08-26 15:15:48 -06:00
Xiang Xiao
42f7c3e927 net/usrsock: Optimize option dependence 2018-08-26 15:05:46 -06:00
ligd
3ae93762b7 net/usrsock/usrsock.h: Fix re-definitions of struct iovec. 2018-08-26 15:03:32 -06:00
Xiang Xiao
31b3dc6c27 net/usrsock: Combine some the duplicated logic. 2018-08-26 15:01:53 -06:00
Gregory Nutt
67193ff5c4 net/usrsock/: Correct of semphore usage issues. 2018-08-26 14:59:04 -06:00
dongjianli
b48804cfc2 net/netdev/netdev_ioctl.c: Add si_ioctl callback and net_ioctl_arglen so usrsock could forward the ioctl to the remote end 2018-08-26 14:54:26 -06:00
Xiang Xiao
3eedabfded net/socket/getsockname.c: Add psock_getsockname function 2018-08-26 14:43:13 -06:00
Xiang Xiao
fd1405d976 netdev_findby_ipv[4|6]addr return netdev_default() as last resort but don't return loopback device if another network device is in the UP state. 2018-08-26 14:40:47 -06:00
zhangyuan7
da1a323f34 net/tcp/tcp_monitor.c: Fix net unlock issue when tcp close. This bug could leave the net locked. 2018-08-26 14:32:06 -06:00
Xiang Xiao
49548da297 net/Kconfig: Uncomment the prompt of NET_SLIP_PKTSIZE 2018-08-26 14:28:35 -06:00
Xiang Xiao
bcf9360e5f Move psock check from [get|set]sockopt to psock_[get|set]sockopt 2018-08-26 14:25:54 -06:00
Xiang Xiao
3f50451046 fs/vfs/fs_fcntl.c, net/socket/net_dupsd.c, and net/socket/net_vfcntl.c: Add file_fcntl, psock_fcntl, and psock_dupsd for use within the kernel. 2018-08-26 13:31:18 -06:00
Gregory Nutt
5d042b41c3 Rename the file devif_loopbackout.c to devif_loopback.c 2018-08-25 08:36:02 -06:00
Gregory Nutt
9bc951a335 Rename devif_loopback_out to devi_loopback 2018-08-25 08:33:21 -06:00
Gregory Nutt
3bf96c8e7e net/arp: Improve arp_find() commit 9774d35010 to reduce the amount and frequency of data copies. 2018-08-25 07:12:21 -06:00
Gregory Nutt
324e1412ef net/arp: Costmetic update to comments. 2018-08-24 17:57:15 -06:00
Gregory Nutt
9774d35010 net/arp: Make the function arp_find() thread-safe. It now returns a copy of the ARP table entry, rather than a potentially unstable reference to the ARP table entry. 2018-08-24 17:36:08 -06:00
Xiang Xiao
0074afa0ac net/netdev: add devif_loopback_out() to check the loopback case where a packet is being sent to itself. Modify the net driver to call this function in this case. This function will simply re-inject the packet back into the network and the network driver will not put anything on the wire. 2018-08-24 09:21:33 -06:00
Xiang Xiao
f74ddd2a04 net/arp: arp_find() now checks if the target IP belongs to one of the local network devices. 2018-08-24 09:04:29 -06:00
Xiang Xiao
e1202d2ed3 Replace all ASSERT with DEBUGASSERT to save the code space 2018-08-24 06:58:30 -06:00
Gregory Nutt
af9fb080e0 fix a spelling error 2018-08-21 10:33:16 -06:00
Gregory Nutt
2e0860e193 net/procfs/netdev_statistics.c: Was not handling the output correct for the case of an IEEE 802.11 device. 2018-08-21 09:08:29 -06:00
Gregory Nutt
4b92f71a82 net/netdev/netdev_register.c: Eliminate a newly introduced warning. 2018-08-14 13:54:10 -06:00
Gregory Nutt
b00a2e5c54 net/: Change order of logic: Select the user-defined interface name format BEFORE determining the minor device number. Otherwise, the devault interface name will be used. 2018-08-14 08:40:06 -06:00
Gregory Nutt
4a724b1f3c net/: Remove all references to CONFIG_NET_USER_DEVFMT. That feature is now unconditionally enabled. This was done because (1) the feature does not require very much additional memory, and (2) it causes confusion in the configuration due to the additional complexity. And network drivers that fail to zero the device structure interface name field (d_ifname) because calling netdev_register() will, however, get a nasty surprise. 2018-08-14 08:01:52 -06:00
Gregory Nutt
8c61c2f31b Remove trailing spaces at the end of lines. 2018-08-13 07:39:38 -06:00
Gregory Nutt
c2267a57a0 Kconfig files: Fix several errors noted by Alex Denisov in Bitbucket issue 115. 2018-08-05 10:48:02 -06:00
Gregory Nutt
056d704cf9 This commit brings in a fragmentary, experimental implementation of NETLINK sockets. There is not too much to that socket support on this initial commit, just the netlink socketer framework. However, I decided to bring it into master because there is a enough that I would not want to lose what is in place. And since it is dependent on CONFIG_EXPERIMENATL, its presence on master should be innocuous. 
Squashed commit of the following:

    net/netlink:  Mark netlink support as EXPERIMENTAL.
    net/netlink/netlink_sockif.c:  Add netlink_getpeername to the socket interface.
    net:  Add getpeeername() support for netlink sockets.
    include/netpacket/netlink.h:  Add a few more definitions and structures used at the NetLink interface.  Still missing many.
    net/netlink:  Add basic framework for Netlink socket support.
    include/:  Add basic Netlink definitions.
 2018-08-03 13:22:36 -06:00
Masayuki Ishikawa
ac5b2ea049 Merged in masayuki2009/nuttx.nuttx/fix_tcp_statistics (pull request #703) 
net/tcp: Remove g_netstats.tcp.syndrop++ from tcp_data_event()

Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>

Approved-by: GregoryN <gnutt@nuttx.org>
 2018-08-03 01:25:53 +00:00
Gregory Nutt
ac5618239c Fix some errors in last commits found in build testing. 2018-07-19 10:30:52 -06:00
Gregory Nutt
788c5bf1dd Forgot to add files before previous commit. 2018-07-19 09:07:20 -06:00
ligd
e840038f2d net/ and include/nuttx/net: Add getpeername() support 2018-07-19 07:16:30 -06:00
Jake Choy
b1daa22040 net/udp/udp_finddev.c: Fix for network byte-order issue when checking if an address is a multicast address. 2018-07-17 11:18:11 -06:00
Alan Carvalho de Assis
283b73edc5 Fix lots of typos in C comments and Kconfig help text 2018-07-08 18:24:45 -06:00
Gregory Nutt
1783d344dc Eliminate some new warnings found in build testing. 2018-07-08 13:44:39 -06:00
Gregory Nutt
6571d4fa37 Update some comments. 2018-07-07 15:23:46 -06:00
Gregory Nutt
427b3b8fcb Squashed commit of the following: 
net/utils:  return from net_breaklock() was being clobbered.
    net/:  Replace all calls to iob_alloc() with calls to net_ioballoc() which will release the network lock, if necessary.
    net/utils, tcp, include/net:  Separate out the special IOB allocation logic and place it in its own function.  Prototype is available in a public header file where it can also be used by network drivers.
    net/utils: net_timedwait() now uses new net_breaklock() and net_restorelock().
 2018-07-07 08:26:13 -06:00
Gregory Nutt
75cc19ebb4 net/tcp: Fix a deadlock condition that can occur when (1) all network logic runs on a single work queue, (1) TCP write buffering is enabled, and (2) we run out of IOBs. In this case, the TCP write buffering logic was blocking on iob_alloc() with the network locked. Since the network was locked, the device driver polls that would provide take the write buffer data and release the IOBs could not execute. This fixes the problem by unlocking the network lock while waiting for the IOBs. 2018-07-06 19:49:05 -06:00
Gregory Nutt
23a8af2069 Trivial update to some comments. 2018-07-06 17:37:26 -06:00
Gregory Nutt
5764d5a529 net/tcp: Back out part of commit d944388888. I see a few places that say that the maximum receive window size is 32,767 (INT16_MAX), but most say that it is 65,535 (UINT16_MAX). 2018-07-05 17:41:01 -06:00
Gregory Nutt
d944388888 configs/viewtool-stm32f107: Add a TCP performance test using apps/examples/tcpblaster. 2018-07-05 16:08:03 -06:00
Gregory Nutt
73c4b1610b net/tcp: Set the default TCP MSS to the value required by RFC 879 and never change it under any circumstance unless the remote host requests a larger MSS via an option the TCP header. 2018-07-05 10:54:45 -06:00