8f7752d956
Networking: Modify how callback structures are stored to avoid another potential use of a stal pointer.
2015-06-03 08:11:57 -06:00
edbb5d0840
networking: Correct a typo introduced with recent networking changes
2015-06-02 15:04:02 -06:00
e22deff1ab
TCP connect. Move the location where the socket is marked as connected in order to avoid a potential race condition
2015-05-31 12:24:24 -06:00
b493dde4cf
TCP networking: In the TCP connection operation, it was trying to setup the network monitor BEFORE the socket was successfully connected. This, of course, has ALWAYS failed because the socket is not yet connected and the TCP state is not yet correct for a connected socket. However, because of other changes net_startmonitor(0 no returns a failure condition that causes worse problems when trying to connect. The fix is to move the logic that starts the network monitor to AFTER the socket has been successfully connected.
2015-05-31 11:35:28 -06:00
a47a0d237c
Costmetic changes to networking logic
2015-05-31 10:29:53 -06:00
33085cb309
Networking: The network device list was protected by a re-entrant semaphore. With the recent change to support network device callback, the network stack needs to access the network device list too. Some drivers, however, run the network stack from the interrupt level -- this is bad but a fact in the current state. Of course,those drivers are unable to take the semaphore and will assert.
...
The solution here is to eliminate the device devices semaphore altogether. This eliminates netdev_semtake() and netdev_semgive() and replaces them with net_lock() and net_unlock() which have larger scope as needed for this purpose.
2015-05-31 08:34:03 -06:00
6687e156e6
Revamp last change. IFF_UP only required when allocating callback, not freeing.
2015-05-30 21:13:58 -06:00
28b46c7bfa
To be valid, a network device must not only exist, but be in the UP state
2015-05-30 12:54:18 -06:00
cc340aaa92
TCP networking: Since the network monitor now allocates a callback structure, it is critical to make sure that the networking monitor is stopped when the socket is closed or any other loss of connection. What was innocuous before would now be a memory leak
2015-05-30 11:49:55 -06:00
8b029fbbee
TCP networking: Hook the network monitor into the device event notification logic
2015-05-30 11:29:47 -06:00
933b95ac80
Forgot to add a file in the last commit
2015-05-30 09:17:46 -06:00
04a661a97c
TCP networking: Add support for network driver events
2015-05-30 09:12:27 -06:00
ab50e9d04d
Networking: Get rid of g_ipv4_allzeroaddr and g_ipv4_alloneaddr. It is more efficient and more intuitive to use INADDR_ANY and INADDR_BROADCAST
2015-05-29 15:16:11 -06:00
34509a9ad5
UDP should use dev=NULL if the address in the lookup is INADDR_ANY
2015-05-29 14:56:44 -06:00
0bdf2d5360
UDP Networking: Misc fixes to get the last changes working + cleanup
2015-05-29 14:32:56 -06:00
fa8b7c19b7
Forgot to add a file in the last commit
2015-05-29 11:03:02 -06:00
f10fdf7ad4
Networking: Add a test to see a device pointer is still valid
2015-05-29 11:01:03 -06:00
2102892e7b
Fix missing file and some compilation errors from the last UDP committ
2015-05-29 10:45:41 -06:00
e672bcebfd
UDP Networking: Add support for device event notification for UDP transfers.
2015-05-29 10:21:06 -06:00
e28a8b1416
Raw sockets: Correct some compile issues; better modularized device-related logic
2015-05-29 08:47:21 -06:00
988fbf0dcc
Merge branch 'master' of ssh://git.code.sf.net/p/nuttx/git into multinic
...
Conflicts:
nuttx/ChangeLog
2015-05-29 08:15:17 -06:00
5337176cdf
Networking: setsockopt() fails when setting timeouts to values less that an 100 msec. That is because the timeout is limited to stops of 1 decisecond and because the conversion of structure timeval was truncating the microsecond remainder. The utility net_timeval2dsec now accespts and option to determin how it handles the remainder: truncate, discarding the remainder, use the remainder to round to the closed decisecond value, or use any non-zero remainder to the next larger whole decisecond value.
...
For the setsockopt() case, it will always use the final option. This means that a time of zero (i.e., not timeout) can only occur if the value zero was explicated provided to setsockopt().
2015-05-29 07:17:15 -06:00
0595d54470
Correct some comments
2015-05-28 19:44:04 -06:00
4747c04c94
Trivial correction to some comments
2015-05-28 19:43:35 -06:00
b5c005d5a4
Merge remote-tracking branch 'origin/master' into multinic
...
Conflicts:
nuttx/ChangeLog
2015-05-28 19:30:48 -06:00
7ec0725533
Missed a couple of function headers
2015-05-28 19:28:41 -06:00
d319b7c6f0
Networking: PKT logic now uses the device event list
2015-05-28 12:40:05 -06:00
e81f279315
Networking: Modify event list handling: Now there are two event lists each device structure: (1) One is for ARP and ICMP data related evetns, the other is for device related events. Callback allocation/free routines no accept a device paramter as well as a list: If the device paramter is added, then the callback goes into both the connection-related liast AND the device event list. Thus each socket type can received both custom data-related events as well as common device related events.
2015-05-28 12:01:38 -06:00
3606d05888
Merge branch 'multinic'
2015-05-28 09:07:18 -06:00
3260db7dbe
Reorder some network events bitset to make the event class clearer
2015-05-28 08:55:36 -06:00
01d176af76
net_startmonitor.c always returned zero. In the case where a socket has already been closed, it correctly handled the disconnetion event but still returned OK. Returning OK causes the callers of net_startmonitor to assume that the connection was okay, undoing the good things that net_startmonitor did and causing the socket to be marked as connected. This behavior was noted by Pelle Windestam.
2015-05-28 08:23:51 -06:00
0e2b6929b4
Add function headers to prototypes in net/socket/socket.h
2015-05-28 07:26:03 -06:00
d6c3ed1c6f
Networking: Fixes an important TCP bug: 16-bit flags was being converted to 8-bits in a few locations, causing loss of status indications
2015-05-27 17:17:42 -06:00
4fd0192a3e
Networking: No that ARP, ICMP, and ICMPv6 reside in the same list we need to make the definitions of ARP_POLL, ICMP_POLL, and ICMPv6_POLL different because there are small differences in the context for each.
2015-05-27 15:04:51 -06:00
390eeaae69
Add function headers to netdev.h header file
2015-05-27 14:34:15 -06:00
2ecff2a432
Add some missing NETDEV_DOWN support for ICMPv6 auto configuration and router support
2015-05-27 12:49:39 -06:00
7d8e8bc188
Missed a place where NETDEV_DOWN needs to be added to notification flags
2015-05-27 12:22:07 -06:00
b94321cfb4
Extend device specific callbacks to ICMPv6 and ARP. Fix some IPv6 compilation errors that have crept in
2015-05-27 11:39:44 -06:00
8ecb84ed89
Improve icmp handling if MULTINIC enabled. From Max Neklyudov
2015-05-27 09:26:00 -06:00
608f812407
Fix bug in IPv6 router from Max Neklyudov.
2015-05-13 07:26:24 -06:00
f20ac5ef93
Add inclusion of assert.h in networking files where needed to avoid compilation errors. From Max Neklyudov.
2015-05-13 07:24:22 -06:00
8cb4e78a1a
Complete implementation of the IPv6 routing logic. From Max Neklyudov.
2015-05-13 07:22:02 -06:00
0f5c35260b
Local sockets: Local stream sockets had problem of double releasing pipes (both server and client attempt release), which causes wrong pipe pair being closed in multi-client case. Solve by adding per connection instance ID to pipe names. From Jussi Kivilinna (2015-05-12).
2015-05-12 07:47:32 -06:00
3c1af2feed
Local sockets: Add poll support for Unix stream sockets. From Jussi Kivilinna.
2015-05-12 07:41:12 -06:00
2fcb93d2ca
Correct some network lock logic: Two error conditions where the network was not being unlocked and one where it was getting unlocked twice. From Jussi Kivilinna.
2015-05-12 07:28:26 -06:00
1f3ee83134
Increase the size of the number of bytes sent from uint16_t to uint32_t in order to avoid TCP errors with long sessions. For exmple:
...
int hello_main(int argc, char *argv[])
{
uint32_t i;
for(i = 0; i < 65536; i++)
{
printf("Hello, World!!\n");
}
printf("press any key!!\n");
if (getchar()=='t')
return 0;
else
return 1;
}
When ran in a Telnet session, the "press any key" is not displayed because the tcp session closed unexpectedly with:
tcp_input: ERROR: conn->sndseq xx, conn->unacked xx"
This is fixed by increasing the width of conn->sent to 32-bits to prevent overflow.
From Rony XLN
2015-05-11 07:14:25 -06:00
d35926ea4a
Fix typo introduced by commit a00f14a60a8ec54fbcf63e44d00f7101ec7cb5a3
2015-05-08 14:19:18 -06:00
00011d4766
Another newly missing semicolon after a debug assertion. Noted by Librae
2015-04-29 15:56:08 -06:00
621eae86b1
Networking: If IPV6 is not enabled, an warning occurs for an unused variable, which is currently used inside the IPV6 block, but declared outside. From ObitalFox.
2015-04-27 11:28:28 -06:00
7a745c3fa4
Use common macro(s) for calculating link layer header length. From Brennan Ashton.
2015-04-13 13:53:13 -06:00
Gregory Nutt