Commit Graph

4649 Commits

Author SHA1 Message Date
goldsimon
f419231dc3 fixed typo in CHANGELOG 2016-11-24 11:31:46 +01:00
Ambroz Bizjak
8ba7363d11 Optimize passing contiguous nocopy buffers to tcp_write
While TCP_OVERSIZE works only when tcp_write() is used with
TCP_WRITE_FLAG_COPY, this new code achieves
similar benefits for the use case that the caller manages their own
send buffers and passes successive chunks of those to tcp_write()
without TCP_WRITE_FLAG_COPY.

In particular, if a buffer is passed to
tcp_write() that is adjacent in memory to the previously passed
buffer, it will be combined into the previous ROM pbuf reference
whenever possible, thus extending that ROM pbuf rather than allocating
a new ROM pbuf.

For the aforementioned use case, the advantages of this code are
twofold:
1) fewer ROM pbufs need to be allocated to send the same data, and,
2) the MAC layer gets outgoing TCP packets with shorter pbuf chains.

Original patch by Ambroz Bizjak <ambrop7@gmail.com>
Edited by David van Moolenbroek <david@minix3.org>
Signed-off-by: goldsimon <goldsimon@gmx.de>
2016-11-24 11:27:34 +01:00
goldsimon
4c8620e03b Added important post-2.0.0 changes to CHANGELOG to keep track of changes for future 2.0.1 release 2016-11-24 11:21:00 +01:00
goldsimon
eb1de78ce1 prepare CHANGELOG for post-2.0.0 2016-11-24 11:12:22 +01:00
Joel Cunningham
5030fa81a0 bug #49684, api_msg: treat non-blocking ERR_MEM as ERR_WOULDBLOCK
This corrects a case in lwip_netconn_do_writemore() where if a
non-blocking socket receives ERR_MEM in a call to tcp_write(), it would
return ERR_MEM, which would result in ENOMEM coming out of the socket
layer

This case can be gracefully handled by returning ERR_WOULDBLOCK since the
socket is already marked as no longer writable and sent_tcp/poll_tcp will
mark the socket as writable again based on available buffer space

This is very similiar to how ERR_MEM is resolved for blocking sockets
2016-11-23 15:03:43 -06:00
Dirk Ziegelmeier
0f87cb92b8 Add note about UDP multicast behavior fix to UPGRADING document 2016-11-23 13:10:16 +01:00
Dirk Ziegelmeier
09547832ba Fix bug #49662: UDP layer should filter incoming multicast datagrams against the bound IP address
Change lwIP UDP API to match socket behavior. Multicast traffic is now only received on a UDP PCB (and therefore on a UDP socket/netconn) when the PCB is bound to IP_ADDR_ANY.
2016-11-23 12:46:35 +01:00
Sylvain Rochet
8c3c96baf7 PPP, L2TP: fix PPPOL2TP_AUTH_SUPPORT == 0 support
Fix compiler warnings on unused parameters and a function signature
mismatch in PPPAPI.
2016-11-22 22:13:24 +01:00
sg
03a9aac157 dns_enqueue(): minor readability improvement: add local variable "age" to store result of subtraction 2016-11-22 21:34:12 +01:00
David van Moolenbroek
68ec20fffc ipv4/ipv6: restrict loopback-destined traffic
Generally speaking, packets with a loopback destination address -
127.0.0.1 for IPv4 and ::1 for IPv6 - should not be accepted on
non-loopback interfaces.  For IPv4, this is implied by RFC 1122
Sec. 3.2.1.3.  For IPv6, it is mandated by RFC 4291 Sec. 2.5.3.
Failure to perform this filtering may have security implications, as
applications that bind sockets to loopback addresses may not expect
that nodes on the local external network be able to produce traffic
that will arrive at such sockets.

With this patch, lwIP drops packets that are sent to a loopback
address but do not originate from the interface that has the loopback
address assigned to it.  This approach works regardless of whether it
is lwIP or the system using it that implements a loopback netif.  The
only exception that must be made is for configurations that enable
netif packet loopback but disable the lwIP loopback netif: in that
case, loopback packets are routed across non-loopback netifs and would
thus be lost by the new filter as well.

For IPv6, loopback-destined packets are also no longer forwarded; the
IPv4 forwarding code already had a check for that.

As a small performance improvement, the IPv6 link-local/loopback
address check is now performed only once per packet rather than
repeatedly for every candidate netif.
2016-11-22 20:51:36 +01:00
Dirk Ziegelmeier
4076b12ee9 Revert "Apply patch #9165: Allowing udp src port to be 0 in cases when we don't care about outgoing port"
This reverts commit 31b0237c50.
2016-11-22 14:35:45 +01:00
David van Moolenbroek
0034abfa45 Always check whether netif_default is NULL
In general, netif_default may be NULL, and various places in the code
already check for this case before attempting to dereference the
netif_default pointer.  Some places do not perform this check though,
and may cause null pointer dereferences if netif_default is not set.
This patch adds NULL checks to those places as well.
2016-11-21 10:16:28 +01:00
Dirk Ziegelmeier
31b0237c50 Apply patch #9165: Allowing udp src port to be 0 in cases when we don't care about outgoing port 2016-11-18 08:13:15 +01:00
Dirk Ziegelmeier
9366c0eaab I decided to keep the "complexity" of handling IPv6 mapped IPv4 addresses out of netconn API.
Only socket API understands this address type now.
2016-11-17 12:41:00 +01:00
Dirk Ziegelmeier
5d5eeca008 Once more: Try fix compile with clang 2016-11-17 09:23:57 +01:00
Dirk Ziegelmeier
4b7e0f50b7 Fix indent in api_lib.c 2016-11-17 09:14:29 +01:00
Dirk Ziegelmeier
44e430ebc2 Fix compile with MSVC 2010 and remove handling for IP6_ADDR_ANY in netconn_connect() - IP6_ADDR_ANY does not make sense in connect() 2016-11-17 09:12:38 +01:00
Dirk Ziegelmeier
180ba72a06 Dual-stack: Use IPv6 mapped IPv4 addresses in receive() and getaddr() only in socket API, not in netconn API.
It is better to present correct IP types in netconn API.
Netconn API now accepts IPv6 mapped IPv4 addresses as well as IPv6 and IPv4 in send(), bind() and connect(), but does NOT map IPv4 to IPv6 mapped IPv4 in getaddr() and receive() functions.
2016-11-17 08:51:07 +01:00
Dirk Ziegelmeier
792224ead0 Try to fix compile error with clang (found by Erik's Travis-CI) 2016-11-17 08:48:55 +01:00
Dirk Ziegelmeier
e8e853f2cb Move ICMP6 protocol constants to include/lwip/prot/icmp6 2016-11-16 23:49:44 +01:00
Dirk Ziegelmeier
80a24c0399 raw, udp, tcp connect() does NOT need to match exact IP type - when PCB is bound to IPADDR_ANY_TYPE, it is OK to connect to IPv4 or IPv6
This should finally implement task #14187: Dual Stack sendto with socket APIs
2016-11-16 23:39:43 +01:00
Dirk Ziegelmeier
d026a3954a Dual-stack fixes in raw/udp/tcp
bind() may change IP type when previous type is IPADDR_TYPE_ANY
connect() IP type must exactly match bind IP type
Use correct IPADDRx_ANY type when calling ip_route()
2016-11-16 23:30:19 +01:00
Dirk Ziegelmeier
1712b06a64 Work on dual-stack netconn
IPv6 netconns are created as IPADDR_TYPE_ANY raw/udp/tcp PCBs internally
bind, connect and sendto now accept IPv6 mapped IPv4 addresses or IPv4 addresses as argument
getaddr and receive functions now return IPv6 mapped IPv4 addresses instead of IPv4 addresses
This behavior is close to BSD socket API
2016-11-16 22:58:38 +01:00
Dirk Ziegelmeier
010f3550b6 Cleanup macro parameter names 2016-11-16 21:37:41 +01:00
Dirk Ziegelmeier
b70ddf7b54 Cleanup unmap_ipv6_mapped_ipv4() macro 2016-11-16 21:37:41 +01:00
Dirk Ziegelmeier
1dd97e7d53 sockets.c, lwip_sendto(): Remove check whether IP address matches socket type. Is checked in lower layers anyway. 2016-11-16 21:37:41 +01:00
Dirk Ziegelmeier
2f37dc0606 api_msg.c: Partly add support for IPv6 mapped IPv4 addresses
- lwip_netconn_do_getaddr(): Convert IPv4 addresses to IPv6 mapped IPv4 addresses
- lwip_netconn_do_send(): Support IPv6 mapped IPv4 addresses
- Not done: connect(), bind()
2016-11-16 21:37:41 +01:00
Dirk Ziegelmeier
c1eb6d8aa4 Minor: Add macros to map/unmap IPv6 mapped IPv4 addresses 2016-11-16 21:37:41 +01:00
Dirk Ziegelmeier
ad17f345e7 mdns: Use strlen to determine string length of strings contained in a struct - padding seems to be applied to these strings, and sizeof() returns the _padded_ size???
Found by compiling with CLANG with address sanitizer enabled
2016-11-16 21:01:15 +01:00
Dirk Ziegelmeier
ee7a2f346c inet.h: Consistently name some inet <-> ipaddr conversion macros
They operate on ip4_addr_t, not on ip_addr_t
This should be clearly visible in their names
2016-11-16 20:30:13 +01:00
Dirk Ziegelmeier
b99b7577fc memp.c: Only check for overflow/underflow if an element could be allocated
In other words: Don't dereference NULL pointers
2016-11-16 20:15:51 +01:00
Joel Cunningham
f79eabd24b bug #49533: start persist timer when unsent seg can't fit in window
This commit returns LwIP to previous behavior where if the next unsent
segment can't be sent due to the current send window, we start the
persist timer. This is done to engage window probing in the case that
the subsequent window update from the receiver is dropped, thus
preventing connection deadlock

This commit refines the previous logic to only target the following case:
 1) Next unsent segment doesn't fit within the send window (not
    congestion) and there is some room in the window
 2) Unacked queue is empty (otherwise data is inflight and the RTO timer
    will take care of any dropped window updates)

See commit d8f090a759 (which removed this
behavior) to reference the old logic. The old logic falsely started the
persit timer when the RTO timer was already running.
2016-11-16 09:12:47 -06:00
Roberto Barbieri Carrera
81a32e9b06 Fixed bug #49610: Sometimes, autoIP fails to reuse the same address
Signed-off-by: goldsimon <goldsimon@gmx.de>
2016-11-16 12:24:57 +01:00
Joel Cunningham
6dc3a2108a WND_SCALE: duplicate define check & doc cleanup
This commit cleans up a duplicate #if check for LWIP_WND_SCALE in init.c
which was already under #if LWIP_WND_SCALE

This commit also improves documentation for TCP_WND in the window scaling
case to communicate TCP_WND is always the calculated (scaled) window value,
not the value reported in the TCP header

Our developers were confused by having to set both the window and scaling
factor and only after studying the usage of TCP_WND throughout the code, was
it determined to be the calculated (scaled) window
2016-11-11 16:06:54 -06:00
Dirk Ziegelmeier
f965034366 Fix inconsistent return value in udp_sendto_if() - ERR_RTE should always be returned when there is no available route 2016-11-11 11:13:26 +01:00
Dirk Ziegelmeier
fdcd8f6faf Fix bug #49578: lwip_socket_drop_registered_memberships API may not work with LWIP_SOCKET_OFFSET 2016-11-11 11:13:25 +01:00
Erik Ekman
b90682dc8b Update email address
I am leaving Verisure at the end of the month.
2016-11-10 11:19:51 +01:00
Dirk Ziegelmeier
b9dc415178 master: Set version to 2.0.1 2016-11-10 09:59:05 +01:00
Axel Lin
d5dd5241e7 dhcp: Use DHCP_MAX_MSG_LEN_MIN_REQUIRED instead of hard-coded value
Signed-off-by: Axel Lin <axel.lin@ingics.com>
2016-11-07 12:36:38 +01:00
Sylvain Rochet
c4eb52dcff PPP: set protocol_list[] const
This array is a constant list of protocols, save some ram space by
adding the const modifier it deserves.
2016-11-06 20:00:31 +01:00
Sylvain Rochet
1e6c4ac017 PPP, IPCP: check that the peer is allowed to use the IP address it wants
This is done in the pppd upstream and was disabled because we don't have
the allowed addresses list required for the auth_ip_addr function.

This is mostly necessary for PPP in server mode to prevent the peer to
use the IP address it wants instead of the one we want, which is
currently allowed.

Rewrite auth_ip_addr in a simple way where we forbid PPP peer to use
loopback net, a multicast address or a reserved class address. Added
to that we consider that PPP in server mode with peer required to
authenticate must provide the peer IP address, reject any IP address
wanted by peer different than the one we wanted. This is actually
an allowed addresses "list" of one entry that follows what is done
in the unused auth_ip_addr function.
2016-11-06 17:39:59 +01:00
Sylvain Rochet
b978d17ca0 PPP, IPCP: remove obvious FIXME
No-op. This is now unused code and it is pretty self explanatory what
int_option do; it checks that passed parameter is an unsigned integer.
2016-11-06 12:54:26 +01:00
Sylvain Rochet
dde55c6c0e PPP, IPCP: fix reset state before reconnecting
Commit 7df5496e7b revealed a regression introduced in commit 5a71509353
which broke IPCP reset state.

ask_for_local was set to 0 if ouraddr initial value is 0, if
ask_for_local was false go->ouraddr was cleared in reset callback,
commit 5a71509353 breaks it by removing this clearing. This regression
was silent because the whole ppp pcb runtime data was cleared before
reconnecting until commit 7df5496e7b which removed this giant clearing.

Fix it by reintroducing ask_for_local boolean value, with proper initial
value following what unused function ip_check_options do.

Fixes: 7df5496e7b ("PPP, rework initial/reconnect cleanup")
Fixes: 5a71509353 ("PPP, CORE, IPCP: removed useless ask_for_local boolean")
2016-11-05 22:07:04 +01:00
Dirk Ziegelmeier
5d22679c67 Use API function instead of accessing struct members directly 2016-11-05 16:14:11 +01:00
Dirk Ziegelmeier
132dafa3fc Correct macro parentheses in tcpip_priv.h 2016-11-05 15:15:08 +01:00
Dirk Ziegelmeier
54d76ffd6d Remove superfluous local variable in raw.c 2016-11-05 15:13:57 +01:00
Dirk Ziegelmeier
b6f32caaad Fix inet_addr_to_ipaddr_p() macro 2016-11-05 10:51:06 +01:00
David van Moolenbroek
1c9e603299 netif: fix reset of IPv6 addresses in netif_add()
Previously, only the state of the first IPv6 address would be reset,
thus possibly keeping other addresses valid (even though zeroed).
2016-11-04 20:18:25 +01:00
Axel Lin
1a7ba24d13 memp: Fix memp_overflow_check_all failure
p needs to point to LWIP_MEM_ALIGN(memp_pools[i]->base) otherwise it will cause
assertion in overflow checking.

Fixes: c838e1ed5b ("Implement possibility to declare private memory pools")
Signed-off-by: Axel Lin <axel.lin@ingics.com>
2016-11-04 20:16:24 +01:00
Joel Cunningham
d4384cfac4 Sockets: check external FD_SETSIZE against number of sockets
This commit adds a compiler check to verify an external FD_SETSIZE has
enough space to store the configured number of sockets
2016-11-03 08:59:19 -05:00