- Count tiebreaking loss in num_conflicts to include them in rate limit detection
- Restart probing using mdns_resp_restart allowing rate limiting for those cases
This ensure rate limiting is well activated during Apple Bonjour Conformance Tests.
Signed-off-by: Simon Goldschmidt <goldsimon@gmx.de>
This allow Apple Bonjour Conformance Test to not fail with the following tests:
- DISTRIBUTED DUPLICATE SUPPRESSION
- MULTIPLE QUESTIONS - DISTRIBUTED DUPLICATE SUPPRESSION
Signed-off-by: Simon Goldschmidt <goldsimon@gmx.de>
TXT records isn't required to be unique in network, so it shouldn't be
included in probe packets.
Additionnaly, when TXT record is present, the Bonjour Conformance Test
from Apple Inc. always fail because generated probe nevert have TXT record.
Signed-off-by: Simon Goldschmidt <goldsimon@gmx.de>
Called with `MDNS_INITIAL_PROBE_DELAY_MS` or `MDNS_PROBE_DELAY_MS` according to
needs.
When `mdns_resp_restart_delay()` called by `mdns_resp_rename_(netif|service)()`
functions, it is assumed this is because a conflict. So we should not use
`MDNS_INITIAL_PROBE_DELAY_MS` because the Bonjour Conformance Test will
complain like this:
```
START (PROBING)
NOTICE 16:40:09.501911: conflicting probe:
smarTrEMotE-f8d0a4.Local.
ERROR 16:40:09.607288: Device did not provide a sufficient time gap between receiving a conflicting probe and reprobing.
ERROR 16:40:09.607333: expected_time_gap=237,actual_time_gap=105
```
Signed-off-by: Simon Goldschmidt <goldsimon@gmx.de>
- Send service slot index to the mdns result function. In case of conflict, the user
will have to remove the service or rename it.
- Break after hostname conflict in order to managed it first, and managed service name
conflict after.
- Provide a function to get the TXT userdata for a service (allowing app to match with
its own data).
Signed-off-by: Simon Goldschmidt <goldsimon@gmx.de>
When more than one service (just 2) need to be probed for conflict, generation
of the probe packet fail because pbuf is too small!
So OUTPACKET_SIZE renamed to MDNS_OUTPUT_PACKET_SIZE and moved to mdns_opts.h
to allow configuration. Default configuration raise it to 1450 to have enough
space when MDNS_MAX_SERVICES > 1 else it remain 512.
Extract from RFC 6762, chapter 17, Multicast DNS Message Size:
The 1987 DNS specification [RFC1035] restricts DNS messages carried
by UDP to no more than 512 bytes (not counting the IP or UDP
headers). For UDP packets carried over the wide-area Internet in
1987, this was appropriate. For link-local multicast packets on
today's networks, there is no reason to retain this restriction.
Given that the packets are by definition link-local, there are no
Path MTU issues to consider.
Multicast DNS messages carried by UDP may be up to the IP MTU of the
physical interface, less the space required for the IP header (20
bytes for IPv4; 40 bytes for IPv6) and the UDP header (8 bytes).
Signed-off-by: Simon Goldschmidt <goldsimon@gmx.de>
sys_arch_mbox_tryfetch() shall return SYS_MBOX_EMPTY or 0 according
to the documentation. Wherever the function is used the return
value is incorrectly compared to SYS_ARCH_TIMEOUT. For now
SYS_MBOX_EMPTY is defined to SYS_ARCH_TIMEOUT so this is not an
issue as long as SYS_MBOX_EMPTY isn't re-defined.
Signed-off-by: Simon Goldschmidt <goldsimon@gmx.de>
As written in RFC5227 in 2.1.1 Probe Details:
A host implementing this specification MUST take precautions to limit
the rate at which it probes for new candidate addresses: if the host
experiences MAX_CONFLICTS or more address conflicts on a given
interface, then the host MUST limit the rate at which it probes for
new addresses on this interface to no more than one attempted new
address per RATE_LIMIT_INTERVAL.
But `acd_restart` restart function check for `acd->num_conflicts > MAX_CONFLICTS`
which allow one more probe than expected.
So this commit change the test to `acd->num_conflicts >= MAX_CONFLICTS`.
Signed-off-by: Simon Goldschmidt <goldsimon@gmx.de>
* Socket functions definitions moved out of the
#define LWIP_SOCKET_EXTERNAL_HEADERS as all users who
set LWIP_SOCKETS to 1 will need them regardless they use
lwip's or external socket headers.
* Include lwip/inet.h in some unit tests and apps
* Since they use htons() and pals.
* test/unit/api/test_sockets.c:
* write() could be declared by external socket headers
* Call lwip_write() instead.
* Code expects fcntl() to return 6
* But O_RDWR could have another value if external
socket headers are present
* Replace 6 by O_RDWR.
* apps/tftp/tftp.c:
* recv() could be declared by external socket headers
* Rename it to tftp_recv()
* Lwip declares msghdr->msg_iovlen as int, but when
using external socket headers, some systems declare
msg_iovlen as size_t or others.
* This patch creates a new type msg_iovlen_t and
expects users to typedef it to the type they need
for their system.
Lwip's struct sockaddr includes sa_len, but some systems
like Linux doesn't have this filed, which produces many
compilation problems when using external headers.
A set of macros has benn added to detect the absence of
sa_len and adapt sockets.c
* LWIP_MARK_TCPIP_THREAD moved to include/lwip/sys.h
* Unix port macro definitions moved to sys_arch.h
* LWIP_MARK_TCPIP_THREAD
* LOCK_TCPIP_CORE
* UNLOCK_TCPIP_CORE
(goldsimon@gmx.de: fixed unix Makefile build and win32 build)
Signed-off-by: Simon Goldschmidt <goldsimon@gmx.de>
In timeouts.c commit 7d1c26cc0c replaced
timeout for AUTOIP with a timeout for ACD, however the value of
LWIP_NUM_SYS_TIMEOUT_INTERNAL was not updated and still counts
LWIP_AUTOIP instead of LWIP_ACD. If user has AUTOIP disabled (or not
explicitly enabled) and DHCP enabled, then ACD gets automatically
enabled too. In this case there will be one timeout too little for lwIP
and first TCP packet received causes an assertion.
Also add LWIP_IPV6_DHCP6 to the value of LWIP_NUM_SYS_TIMEOUT_INTERNAL,
as it was also not accounted for.
This reuses the member 'int socket' by making it a union containing
both int and void pointer.
See bug #56593.
Signed-off-by: Simon Goldschmidt <goldsimon@gmx.de>
Suggested-by: Wilfred <wilfrednilsen@hotmail.com>
The err field is removed from struct lwip_sock since commit e0a2472706
("netconn/sockets: remove fatal error handling, fix asynchronous error handling, ensure data before RST can be received")
sock_set_errno() simply calls set_errno() now, so use set_errno() instead.
Signed-off-by: Axel Lin <axel.lin@ingics.com>
Acked-by: Dirk Ziegelmeier <dziegelmeier@de.pepperl-fuchs.com>
If client reception buffer is bigger than the first frame we receive, the first packet test
will always fail for the second one if it is shorter the the diffence between reception
buffer size and first frame length.
For example, if we receive a PUBLISH message with length = 1517 (payload len = 1514 +
header len = 3), this result in total message length of 1517.
altcp_tls will send MQTT client frame up to 1516 bytes max. This result to PUBLISH
message splitted in two frame: first is 1516 bytes, the second of 1 bytes.
If MQTT_VAR_HEADER_BUFFER_LEN is 1520 (1516 + 4 bytes for stored fixed header), the
second frame of 1 bytes is considered as first publish frame because
client->msg_idx (1517) < MQTT_VAR_HEADER_BUFFER_LEN (1520).
This result in disconnection AND application callback never called for the end of the
payload.
The fix will check `(client->msg_idx - (fixed_hdr_len + length)) == 0` which can be
only true for the first frame of a message.
Below logs showing the bug:
```
April 3rd 2019, 23:14:05.459 lwip_dbg mqtt_parse_incoming: Remaining length after fixed header: 1514
April 3rd 2019, 23:14:05.460 lwip_dbg mqtt_parse_incoming: msg_idx: 1516, cpy_len: 1513, remaining 1
April 3rd 2019, 23:14:05.460 lwip_dbg mqtt_incomming_publish: Received message with QoS 1 at topic: v2/inte...
April 3rd 2019, 23:14:05.461 lwip_dbg mqtt_parse_incoming: Remaining length after fixed header: 1514
April 3rd 2019, 23:14:05.461 lwip_dbg mqtt_parse_incoming: msg_idx: 1517, cpy_len: 1, remaining 0
April 3rd 2019, 23:14:05.461 lwip_dbg mqtt_message_received: Received short PUBLISH packet
```
