goldsimon
6d95a34971
icmp ping response: fix invalid checksum (and possible assertion failure) when ip header contains options (is it correct that we mirror back all options)
2016-06-17 10:07:49 +02:00
goldsimon
61e067b98a
Check for minimum IPv4 header length in rx packets
2016-06-17 09:36:14 +02:00
goldsimon
fda778f6b6
ip6_reass: don't crash if frag header isn't the first (or not in the first pbuf)
2016-06-16 13:51:38 +02:00
Erik Ekman
d98e25a783
nd6: Handle incorrect ICMP option length in RA
...
Make sure ICMPv6 options have a valid length before
parsing them.
Found with afl-fuzz.
2016-06-14 15:06:49 +02:00
sg
4e241fbf14
some minor improvements to the tcp options checking code
2016-06-13 20:41:58 +02:00
Henrik Persson
b9a2ee8aaa
[PATCH] Drop instead of ASSERT in tcp_input header parsing
...
Since allowing input validation to trip the ASSERT handler is bad,
let's just drop the packets instead if validation fails.
Signed-off-by: sg <goldsimon@gmx.de>
2016-06-13 20:08:43 +02:00
Dirk Ziegelmeier
5e7b343d52
Fix dns_getserver to return a pointer, not a value
2016-06-13 09:00:46 +02:00
Dirk Ziegelmeier
6be7e221a5
Fix a few doxygen warnings
2016-05-24 22:29:18 +02:00
sg
631c458c55
Fixed invalid DEBUGF level SERIOUS when pbuf_header fails (it returns '1': this is not serious!)
2016-05-23 22:06:02 +02:00
Dirk Ziegelmeier
90a656ed78
Fix several doxygen errors all over the code
2016-05-23 21:18:16 +02:00
Jan Breuer
ebd103775d
ipv6_addr: Render last zero in ip6addr_ntoa_r if not in first empty block
2016-05-23 09:54:54 +02:00
goldsimon
bd177ff38f
added comment, fixed coding style
2016-05-23 09:47:14 +02:00
Jan Breuer
ac21a5f370
nd6: add cleanup function
2016-05-23 09:44:27 +02:00
Axel Lin
8c0f620d83
dns: Use LWIP_ARRAYSIZE instead of open coded
...
Signed-off-by: Axel Lin <axel.lin@ingics.com>
2016-05-11 12:20:15 +02:00
Axel Lin
7e9377a9c7
dhcp: Use LWIP_ARRAYSIZE to get the number of array entries
...
dhcp_discover_request_options is u8_t array, so the result is the same.
But use LWIP_ARRAYSIZE to get the number of array entries is better
because it works for all types.
Signed-off-by: Axel Lin <axel.lin@ingics.com>
2016-05-11 12:15:12 +02:00
Sylvain Rochet
1dcd5d31d7
pbuf: fix undefined PBUF_POOL_FREE_OOSEQ if TCP_QUEUE_OOSEQ is disabled
...
lwip/src/core/timers.c: In function ‘sys_check_timeouts’:
lwip/src/core/timers.c:328:5: error: "PBUF_POOL_FREE_OOSEQ" is not defined [-Werror=undef]
#if PBUF_POOL_FREE_OOSEQ
Fix it by declaring an empty PBUF_CHECK_FREE_OOSEQ() function if feature is
not enabled.
2016-05-08 03:09:51 +02:00
Dirk Ziegelmeier
49badf39a1
Fix bug #47805 : dhcp_recv can dereference NULL pointer
2016-04-29 11:16:33 +02:00
Dirk Ziegelmeier
2e05856f90
Fix bug #47641 : lwIP would accept IPv6 source addresses from the network that are mapped IPv4 adresses
2016-04-27 21:35:34 +02:00
goldsimon
c981b726f8
fixed bug #47601 (wrong NS request at DAD)
2016-04-27 10:19:13 +02:00
goldsimon
afd9ee9147
fixed bug #47743 (Closing listening tcp pcb is not posible without assert)
2016-04-26 13:27:17 +02:00
Dirk Ziegelmeier
bb908366a5
More PPP cleanups
2016-04-25 20:08:29 +02:00
Dirk Ziegelmeier
73d8f14ebd
Fix compile when PPP is not enabled
2016-04-25 20:08:29 +02:00
Dirk Ziegelmeier
ffa340a68c
task #13907 : PPP cleanups: Move PPP mempools out of lwIP core to PPP code
2016-04-25 20:08:29 +02:00
Dirk Ziegelmeier
4d1a2e6699
Fix projects without PPP do not compile any more
2016-04-22 06:24:01 +02:00
Sylvain Rochet
15fbfb7363
PPP, remove useless ppp_init function
...
Simon says:
ppp_init() does not seem to be used. The only thing it does is calling
magic_init(), which is not required because it its called again later
from ppp_input().
Also, the time from startup is rather constant, so calling sys_jiffies()
from ppp_init() does not create a random number.
2016-04-21 23:25:48 +02:00
Joel Cunningham
8ac1b4825a
udp: use netif addr as src for multicast sockets
...
Sockets bound to a multicast address could not transmit multicast
packets because the pcb local address did not match the netif address
even if the outgoing netif was resolved correctly.
To correct the issue, pcbs with a multicast local address will use
the outgoing netif address as the source address in IPv4.
2016-04-18 16:36:13 -04:00
Dirk Ziegelmeier
309e072238
Create new IP_IS_V4 macros and use them at instead of !IP_IS_V6 - since we now have an IPADDR_ANY_TYPE, just checking for !V6 does not mean it is V4
2016-04-06 22:31:31 +02:00
sg
ca1b8288de
prepare for overriding current timeout implementation: all stack-internal caclic timers are avaliable in the lwip_cyclic_timers array
2016-04-05 23:01:57 +02:00
sg
8c7f513f4d
patch #8358 : allow more combinations of listening PCB for IPv6
2016-04-05 22:04:19 +02:00
sg
0731eda0b3
make autoip_supplied_address() take a const pointer, too
2016-04-05 20:38:11 +02:00
Dirk Ziegelmeier
12a4cae404
Apply [patch #8969 ] from Freddie Chopin: Constify netif in dhcp_supplied_address()
2016-04-05 13:18:27 +02:00
sg
27f03798b9
Fixed possible problems with tcp_backlog_delayed/tcp_backlog_accepted
2016-03-25 16:19:39 +01:00
goldsimon
5b0a47ca87
removed unused return values of tcp_listen_input & tcp_timewait_input; fixed unused variable warning
2016-03-24 15:06:31 +01:00
sg
7721b20179
call accept-callback with ERR_MEM when allocating a pcb fails on passive open to inform the application about this error; ATTENTION: applications have to handle NULL pcb in accept callback!
2016-03-23 21:57:38 +01:00
sg
c6b742812d
fixed bug #38203 : DHCP options are not recorded in all DHCP ack messages (patch by Florent Matignon)
2016-03-23 21:06:52 +01:00
Axel Lin
449fc98a55
tcp_alloc(): Remove superfluous initialization for pcb->keep_cnt_sent
...
It's already done by memset(0). This was missed in commit 5794ac2340
"tcp_alloc(): remove explicit =0 already done by memset(0)", fix it.
Signed-off-by: Axel Lin <axel.lin@ingics.com>
2016-03-23 14:26:32 +01:00
Dirk Ziegelmeier
f97bf85fd6
udp.c: Allow udp_connect() on a dual-stack UDP PCB (type IPADDR_TYPE_ANY). This may be useful in some applications - no need to keep separate storage for target IP.
2016-03-22 15:12:16 +01:00
sg
dd80759bb9
tcp: changed accept handling to be done internally: the application does not have to call tcp_accepted() any more. Instead, when delaying accept (e.g. sockets do), call tcp_backlog_delayed()/tcp_backlog_accepted() (fixes bug #46696 )
2016-03-22 07:30:44 +01:00
sg
a1c78ea7bf
ignore dns response parsing errors, only abort resolving for correct responses or error responses from correct server (bug #47459 )
2016-03-22 07:22:17 +01:00
sg
5794ac2340
tcp_alloc(): remove explicit =0 already done by memset(0)
2016-03-18 22:05:17 +01:00
sg
44e1a2d8e2
define tcp_backlog_set() as dummy-define when backlog feature is disable
2016-03-18 20:05:04 +01:00
sg
852d5b9c80
Optimize tcp_output runtime by not calling ip_route() for every segment sent
2016-03-16 22:31:33 +01:00
sg
b4d7238eb8
tcp_output_segment: don't count retransmitted segments in mib2.tcpoutsegs by detecting p->payload != tcphdr pointer; don't try to retransmit segments where p->ref != 1 (as it is invalid to mess up p->len/p->payload when we don't have exclusive access to the pbuf)
2016-03-16 22:22:20 +01:00
sg
5f8b5cbb20
reset rto timer on fast retransmission (see task #13757 , patch by Joel Cunningham)
2016-03-16 21:49:37 +01:00
sg
8e8571da6a
fixed bug #46384 Segment size calculation bug with MSS != TCP_MSS
2016-03-16 16:56:15 +01:00
Dirk Ziegelmeier
f3b7bca3cf
Fix bug #47370 : Port DNS client to IPv6
2016-03-09 19:25:06 +01:00
Dirk Ziegelmeier
a212210c35
Add missing #include in memp.c
2016-03-08 19:53:01 +01:00
sg
5cf802eda0
dns_gethostbyname: return error code if no valid server is set instead of calling the 'found' callback in this case (fixes bug #46887 )
2016-03-07 19:47:39 +01:00
sg
f5fef6eebc
fixed bug #45353 : IPv6 router / neighbor relationship is not always consistent - patch by Stian Sebastian Skjelstad
2016-03-06 14:48:41 +01:00
sg
b00a17059a
fixed icmp_send_response with LWIP_HOOK_IP4_ROUTE_SRC enabled
2016-03-05 15:57:40 +01:00
sg
d2bb569ee4
bug #46094 : moved IPv4 routing hook functions down so that they implement actual routing (source-based IPv4 routing is not affected by this change)
2016-03-05 15:52:39 +01:00
sg
e0ab8c581d
fixed bug #46093 : Move IP6_ROUTE() hook to after LL processing
2016-03-05 14:11:24 +01:00
sg
6b405eccd0
Tried to fix bug #47270 by accepting all ipv4 multicast packets (at IP level) when LWIP_IGMP is disabled
2016-03-04 21:57:47 +01:00
Dirk Ziegelmeier
9a4c66006e
Fix compile error in raw.c (possible unused arg)
2016-03-03 22:55:42 +01:00
Dirk Ziegelmeier
1dea649cff
Fix comment in tcp.c
2016-03-03 22:48:10 +01:00
Dirk Ziegelmeier
7396d8818b
Implement dual-stack in RAW API in the same way as in UDP and TCP
2016-03-03 22:47:27 +01:00
Dirk Ziegelmeier
2d80af1dcb
Minor code cleanup in udp.c
2016-03-03 22:44:01 +01:00
Dirk Ziegelmeier
5358380c5f
Apply patch from Derek Pravala: Fix segfault in ip6_route when there is no default netif
2016-03-03 16:12:20 +01:00
Dirk Ziegelmeier
e89b48d23f
raw.c: Avoid NULL pointer dereference
2016-03-02 23:44:51 +01:00
Dirk Ziegelmeier
f9ce31f98b
TCP: Implement dual stack support using new IP_ANY_TYPE introduced at UDP
2016-03-02 23:14:33 +01:00
Dirk Ziegelmeier
f3132c5290
udp: Fix possible NULL pointer dereference, update comments
2016-03-02 23:10:43 +01:00
Dirk Ziegelmeier
9885d5d9f5
Apply patch from Simon to fix buffer overflow handling tcp_in.c which was discovered by Coverity
2016-02-29 23:34:52 +01:00
Dirk Ziegelmeier
8e5663a32f
Minor: Avoid ip_input() call indirection for when only IPv4 or IPv6 are enabled
2016-02-29 22:17:42 +01:00
Dirk Ziegelmeier
6aed6e659f
Rework DHCP PCB handling: Old code registered one UDP PCB per netif where DHCP was active and there was a special case in udp_input() for this. New implementation uses one PCB for all netifs and removes special case in udp_input().
...
The old approach called udp_bind() on each of the PCBs, which puts them into udp_pcbs list. The PCBs were iterated on all non-DHCP udp_inputs() with no effect.
My cleanup removes the special handling in udp.c, and uses only one DHCP UDP PCB to catch all DHCP messages from all netifs. The dhcp_recv function then checks whether ip_current_input_netif() has DHCP enabled - if not, the message is ignored. The PCB is only created/registered when one or more PCBs have DHCP enabled.
2016-02-29 19:31:19 +01:00
Dirk Ziegelmeier
70f3e5ed05
Cleanup: Create new file ip.c and move dual-stack code from ipv4 and ipv6 folder in there
2016-02-26 23:04:51 +01:00
Dirk Ziegelmeier
d5778bad2f
No need for dual-stack code in ip4.c and icmp.c -> use ip4 only versions
2016-02-26 22:50:20 +01:00
Dirk Ziegelmeier
a5214abd04
udp: udp_send and udp_send_chksum: Check for PCB NULL pointer and PCB must not be the new IP_ANY_TYPE, but a real IPv4 and IPv6 address
2016-02-26 09:20:50 +01:00
goldsimon
bf8bcfafe2
revert accidentally committed comment from a649a3eb50
2016-02-26 08:59:50 +01:00
Joel Cunningham
8cb64698aa
Allow backlog to be updated
...
This commit adds support to the sockets and netconn layer to update the
backlog by calling listen when the netconn is already in the listen state.
When backlog is not enabled, the call returns successfully
This commit also introduces a macro for setting the backlog value that
prevents a 0 sized (invalid) backlog
2016-02-25 15:19:52 -06:00
sg
a649a3eb50
dual-stack fix: moved common definitions/code for ethernet used by etharp and ethip6 to new files ethernet.h/.c
2016-02-25 22:10:28 +01:00
Dirk Ziegelmeier
fd891081c4
minor: Cleanups in raw/tcp/udp code by using macros and reducing #ifdefs
2016-02-24 23:04:25 +01:00
Dirk Ziegelmeier
953b7bdd59
Implement UDP dual-stack PCB support
...
Create special IP address type "IPADDR_TYPE_ANY" for it.
SNMP uses new feature in non-netconn mode.
TODO: Same for TCP & RAW, adapt NETCONN to use this feature
2016-02-24 22:37:01 +01:00
sg
1dde3d6e56
Fixed bug #46524 : "Device as dhcp server does not work after upgrade to lwip-1.4.1"
2016-02-24 22:13:58 +01:00
sg
ec49b68b38
minor: udp_input/local_match: coding style, added comments
2016-02-24 22:04:59 +01:00
Dirk Ziegelmeier
ac4b089b18
udp.c: Further udp_input simplifications
...
- Move local PCB matching code in a function that can be reused in SO_REUSE && SO_REUSE_RXTOALL case.
- Some checks have been written in the dual-stack version and then repeated with the ipv6-only version. Example:
IPv6 only: ip6_addr_ismulticast(ip6_current_dest_addr())
IPv4 AND IPv6: ip_addr_ismulticast(ip_current_dest_addr())
2016-02-24 19:57:20 +01:00
Dirk Ziegelmeier
005c196ed8
udp.c: Simplify and cleanup input PCB matching
2016-02-24 19:13:30 +01:00
Dirk Ziegelmeier
17fad79f71
Two compile fixes after eliminating isipv6 member in PCBs
2016-02-22 17:07:17 +01:00
Dirk Ziegelmeier
c805843e64
Change signature of ip_output, ip_output_if, ip_output_if_src, ip_output_hinted, ip_route, ip_netif_get_local_ip, ip_route_get_local_ip and tcp_eff_send_mss not to take an isipv6 parameter. Use the IP version of the destination address instead.
2016-02-22 13:38:40 +01:00
Dirk Ziegelmeier
880f1008d6
Together with Simon: Eliminate last instances of isipv6 member in pcbs;
...
Don't allow NULL pointers (IPv4 ANY) in connect() and sendto() functions as destinations since this does not make sense.
2016-02-22 13:03:51 +01:00
Dirk Ziegelmeier
be9cd800b0
Review with Simon: Init TCP/UDP/RAW PCBs with correct IP types in _new functions to make my changes work correct
2016-02-22 11:16:13 +01:00
Dirk Ziegelmeier
0106cf7ae0
Review with Simon: Remove dead code (since 2003)
2016-02-22 10:18:19 +01:00
Dirk Ziegelmeier
50303eea23
Reduce PCB_ISIPV6 macro usage some more
2016-02-21 19:59:08 +01:00
Dirk Ziegelmeier
339e82d7aa
Eliminate IP_PCB_IPVER_EQ macro
2016-02-21 19:45:51 +01:00
Dirk Ziegelmeier
5a25652c21
Eliminate IP_PCB_IPVER_INPUT_MATCH macro
2016-02-21 19:21:36 +01:00
Dirk Ziegelmeier
8f675c37fc
Reduce usage of PCB IP version flag in UDP code. Maybe we can figure out a good way for dual-stack UDP when we manage to remove the flag (nearly) entirely from the code.
2016-02-20 20:54:58 +01:00
Dirk Ziegelmeier
5809b01388
Reduce usage of PCB IP version flag in raw and tcp code. Maybe we can figure out a good way for dual-stack UDP when we manage to remove the flag (nearly) entirely from the code.
2016-02-20 20:40:04 +01:00
Dirk Ziegelmeier
10edba936b
Fix tcp pcb access after it was freed in tcp.c - found by Coverity
2016-02-19 23:28:09 +01:00
goldsimon
e4cac595c2
Check TCP_SNDLOWAT for too high value (near u16_t overflow) - fixes bug #46948
2016-02-18 09:36:55 +01:00
Dirk Ziegelmeier
24bc77a935
Found some struct instances that can be const
2016-02-16 22:31:59 +01:00
Dirk Ziegelmeier
273d609bac
mem.c: Fix unintended sign extension (found by Coverity)
...
sign_extension: Suspicious implicit sign extension: count with type unsigned short (16 bits, unsigned) is promoted in count * size to type int (32 bits, signed), then sign-extended to type unsigned long (64 bits, unsigned). If count * size is greater than 0x7FFFFFFF, the upper bits of the result will all be 1.
2016-02-10 21:21:08 +01:00
Dirk Ziegelmeier
288b4564e4
DNS: Fix potential array out of bounds access. Variable nanswers may contain bogus values. (found by Coverity)
2016-02-10 20:41:31 +01:00
Dirk Ziegelmeier
1c1587eef6
ICMP: Fix responses were always sent to default netif because of uninitialized source IP (found by Coverity)
2016-02-10 20:35:58 +01:00
Dirk Ziegelmeier
11b810721a
ICMP: Fix unused value in icmp_input (found by Coverity)
2016-02-10 20:34:44 +01:00
Dirk Ziegelmeier
222155d7e2
pbuf: Make pbuf_skip function public, so it can be used in SNMP code
2016-02-03 08:37:04 +01:00
Axel Lin
f411a34af5
stats: Add proper #if guard to prevent unused-variable warning for memp_names
...
The usages of memp_names are under either LWIP_DEBUG in stats_init,
or under LWIP_STATS_DISPLAY in stats_display_memp.
Fix below build warning:
lwip/src/core/stats.c:53:21: error: 'memp_names' defined but not used [-Werror=unused-variable]
static const char * memp_names[] = {
^
cc1: all warnings being treated as errors
Reported-by: David Fernandez
Fixes: 2f2a75a6d9fd ("stats: Move memp_names table out of stats_init/stats_display_memp functions")
Signed-off-by: Axel Lin <axel.lin@ingics.com>
2016-01-31 13:48:38 +01:00
Axel Lin
515a4cad73
Clear addr_hint pointer after calling ip_output_if() in tcp_zero_window_probe()
...
This was missed in commit aa0e41c389
("task #12178 : hardware checksum capabilities can be configured per netif"),
fix it.
Signed-off-by: Axel Lin <axel.lin@ingics.com>
2016-01-29 09:21:58 +01:00
Stephan Linz
2c31beb6e7
IPv4: fix some meaningless debug messages
...
Signed-off-by: Stephan Linz <linz@li-pro.net>
2016-01-18 21:34:13 +01:00
Axel Lin
fe2cfe2dba
raw: Fix build error
...
IP6_HLEN is only defined when LWIP_IPV6, IP_HLEN is only defined when LWIP_IPV4.
This fixes build error in !LWIP_IPV4 || !LWIP_IPV6 cases.
Fixes: f2c7e9c939
("raw: Remove unnecessary #if guard around PCB_ISIPV6() calls")
Reported-by: Erik Ekman
Signed-off-by: Axel Lin <axel.lin@ingics.com>
2016-01-18 16:02:01 +01:00
Stephan Linz
80375e68dd
ip_frag: fix compiler warnings, unused variable
...
In case of IP_REASS_FREE_OLDEST==0, the argument clen of
ip_reass_enqueue_new_datagram() will not used and leads
the compiler to a warning.
Signed-off-by: Stephan Linz <linz@li-pro.net>
2016-01-18 11:55:54 +01:00
Axel Lin
f2c7e9c939
raw: Remove unnecessary #if guard around PCB_ISIPV6() calls
...
PCB_ISIPV6() macro is well defined for all cases (LWIP_IPV4 && LWIP_IPV6,
LWIP_IPV4 only and LWIP_IPV6 only), thus remove the unnecessary #if guard
around PCB_ISIPV6() calls.
Signed-off-by: Axel Lin <axel.lin@ingics.com>
2016-01-18 11:55:06 +01:00