Lines Matching refs:is
117 * ppp.c: fixed pbuf leak when PPP session is aborted through pppSigHUP()
125 * msg_in.c, msg_out.c: fixed bug #35536 SNMP: error too big response is malformed
161 netif is brought down
186 * sockets.h: fixed bug #34580 fcntl() is missing in LWIP_COMPAT_SOCKETS
192 * tcp_in.c, tcp_out.c: fixed bug #34517 (persist timer is started although no
193 zero window is received) by starting the persist timer when a zero window is
198 * def.h, timers.c: fixed bug #34541: LWIP_U32_DIFF is unnecessarily complex
201 * sockets.c, api_lib.c: fixed bug #34540: compiler error when CORE_LOCKING is
224 * mem.h: fixed bug #34377 MEM_SIZE_F is not defined if MEM_LIBC_MALLOC==1
246 * udp.c: fixed bug #34072: UDP broadcast is received from wrong UDP pcb if
250 * tcp_in.c: fixed bug #33952 PUSH flag in incoming packet is lost when packet
251 is aggregated and sent to application
254 * opt.h: fixed bug #31809 LWIP_EVENT_API in opts.h is inconsistent compared
277 * tcp_out.c: fixed bug #33962 TF_FIN not always set after FIN is sent. (This
286 * pbuf.c/.h, timers.c: freeing ooseq pbufs when the pbuf pool is empty implemented
292 sending an ARP request when an ARP entry is used in the last minute before
332 link-layer-addressed UDP traffic to be received while a netif is down (just
337 version is now only used in protocol headers. Added global storage for
401 event_callback is only executed when select is waiting.
419 allocated larger than needed and the space is used for later calls to
483 for allocating memory when getaddrinfo() is called.
543 * sys_arch.txt: sys_arch_timeouts() is not needed any more.
553 * tcp.c: Fixed bug #32926 (TCP_RMV(&tcp_bound_pcbs) is called on unbound tcp
557 * ppp.c: Fixed bug #32280 (ppp: a pbuf is freed twice)
565 is present never times out) by starting retransmission timer before checking
570 calling sio_read_abort() if the file descriptor is valid.
578 * sockets.c: fixed bug #32769 (ESHUTDOWN is linux-specific) by fixing
584 connection has been aborted by err_tcp (since this is not a normal closing
602 * err.h: Fixed ERR_IS_FATAL so that ERR_WOULDBLOCK is not fatal.
605 * api.h, api_lib.c, api_msg.c, sockets.c: netconn.recv_avail is only used for
645 * sockets.h: Fixed bug #31385: sizeof(struct sockaddr) is 30 but should be 16
736 triggered by bug #29345; don't allocate acceptmbox if LWIP_TCP is disabled
777 * api_msg.c: Correctly set TCP_WRITE_FLAG_MORE when netconn_write is split
804 reception and is not really necessary
814 (NETIF_FLAG_ETHERNET) that tells the stack the device is an ethernet
1011 * pbuf.c: Fixed bug #28064: pbuf_alloc(PBUF_POOL) is not thread-safe by
1012 queueing a call into tcpip_thread to free ooseq-bufs if the pool is empty
1049 pcb->recv is NULL to keep rcv_wnd correct)
1075 valid DHCP packets -> allow 0.0.0.0 as source address when LWIP_DHCP is
1096 Fixes bug #27338: sys_stats is defined when NO_SYS = 1
1109 * netdb.c, dns.c: bug #26657: DNS, if host name is "localhost", result
1110 is error.
1148 bigger malloc pool if one is empty (only usable with MEM_USE_POOLS).
1177 is otherwise used)
1181 is only used by UDPLITE at present, so conditionalise it.
1186 LWIP_AUTOIP_CREATE_SEED_ADDR is overridden.
1214 loopback traffic is supported on all netifs (all local IPs).
1242 * tcp.h: BUG26879: set ret value in TCP_EVENT macros when function is not set
1269 * dhcp.c, netbios.c: Changes if IP_SOF_BROADCAST is enabled.
1272 * tcp.h: bug #26349: Nagle algorithm doesn't send although segment is full
1280 * msg_in.c: fixed bug #25636: SNMPSET value is ignored for integer fields
1317 pcb->err is called, not pcb->connected (with an error code).
1346 is big enough in dhcp_start
1398 * dhcp.c: fixed bug #25345 (DHCPDECLINE is sent with "Maximum message size"
1409 * tcp.c: fixed bug #25467: Listen backlog is not reset on timeout in SYN_RCVD:
1410 Accepts_pending is decrease on a corresponding listen pcb when a connection
1411 in state SYN_RCVD is close.
1437 padding is included.
1459 * tcp_in.c: Fix for bug #23693 as suggested by Art R. Ensure cseg is unused
1473 passing as function pointers when MEM_LIBC_MALLOC is defined.
1526 "Add return value to sys_mbox_post". tcpip_callback is always defined as
1538 indicate the number of pointers query by the mailbox. There is three defines
1542 The new sys_mbox_trypost function return a value to know if the mailbox is
1543 full or if the message is posted. Take a look to sys_arch.txt for more details.
1544 This new function is used in tcpip_input (so, can be called in an interrupt
1545 context since the function is not blocking), and in recv_udp and recv_raw.
1550 "backlog" parameter in an u8_t, 0 is interpreted as "smallest queue", add
1587 * ip.c: ip_input: check if a packet is for inp first before checking all other
1593 pcbs so that it is found faster next time. Same for RAW pcbs that have eaten
1615 copy dns payload to parse the response), return an error if there is no place
1631 requests with RAW api interface. Initialization is done in lwip_init() with
1632 build time options. DNS timer is added in tcpip_thread context. DHCP can set
1635 some traces with LWIP_DEBUGF. Sanity check have been added. There is a "todo"
1641 for ifOperStatus if LWIP_NETIF_LINK_CALLBACK is defined.
1652 it. This is a first release which have to be improve for TCP. Note it used the
1653 netconn::recv_avail which need to be more "thread-safe" (note there is already
1658 Integrate "patch #6250 : MSG_MORE flag for send". MSG_MORE is used at socket api
1661 Note that previous "copy" parameter for "write" APIs is now called "apiflags".
1677 simultaneously and no static reassembly buffer is needed.
1703 decide to call snmp_add_sysuptime(100) each 1000ms (which is bigger "step", but
1706 This one is undefined by default in mib2.c. SNMP_GET_SYSUPTIME is called inside
1708 when it's queried (any direct call to "sysuptime" is changed by a call to
1713 and others without it, there is a new NETIF_FLAG_IGMP flag to set in netif->flags
1714 if you want IGMP on an interface. igmp_stop() is now called inside netif_remove().
1715 igmp_report_groups() is now called inside netif_set_link_up() (need to have
1716 LWIP_NETIF_LINK_CALLBACK=1) to resend reports once the link is up (avoid to wait
1749 how many times a group is used on an netif, and free it when all applications
1763 from api/api_lib". Now netbuf API is independant of netconn, and can be used
1773 igmp_start is call inside netif_add. Now, igmp initialization is in the same
1781 and add init.c in their makefiles. Note that lwip_init() function is called
1783 disabled when matching options are disabled. Also note that their is new options
1792 * netif.h, netif.c: Integrate "patch #6163 : Function to check if link layer is up".
1793 Add a netif_is_link_up() function if LWIP_NETIF_LINK_CALLBACK option is set.
1801 ETHARP_TCPIP_ETHINPUT, now, only "ethinput" code is supported, even if the
1802 name is tcpip_input (we keep the name of 1.2.0 function).
1887 of 6 to avoid loading netif->hwaddr_len every time (since this file is only
1938 elements to overflow. This is achieved by adding some bytes before and after
1942 or memp_free() is called (slower but more helpful).
1970 to be overriden in case the C-library malloc implementation is not protected
1987 sys_arch_mbox_tryfetch function to get a message if one is there, otherwise return
2006 define a hostname in netif struct (this is just a pointer, so, you can use a hardcoded
2012 * netif.h, netif.c: A new NETIF_FLAG_ETHARP flag is defined in netif.h, to allow to
2013 initialize a network interface's flag with. It tell this interface is an ethernet
2015 Support for IPv4" section 4.6) when interface is "up" with netif_set_up().
2019 time if you only use PPP or SLIP. The default is enable. Note we don't have to call
2021 is done in tcpip_init function.
2088 like this is a bit slow and that using (lwip)getpeername is the
2089 good lwip way to do it (so, using recv is faster on tcp sockets).
2124 is not protected" by using new macros for interlocked access to modify/test
2148 netconn, this recvmbox is later freed and a new mbox is allocated for acceptmbox.
2149 This is a fix for thread-safety and allocates all items needed for a netconn
2150 when the netconn is created.
2154 netifs): if LWIP_DHCP is enabled, UDP packets to DHCP_CLIENT_PORT are passed
2159 * ip.c: fixed bug #21643 (udp_send/raw_send don't fail if netif is down) by
2182 of the netconn_peer and netconn_addr processing is done inside tcpip_thread
2190 * tcp.c, tcp_in.c: Fixed bug #21494: The send mss (pcb->mss) is set to 536 (or
2191 TCP_MSS if that is smaller) as long as no MSS option is received from the
2196 is now based on TCP_MSS instead of pcb->mss (on passive open now effectively
2234 it as long as "lwip/opt.h" is included before (this one include "lwip/debug.h" which
2267 This is mainly on using lookup/lookfor, and some coding styles...
2275 (tcp_output is called again later from tcp timers).
2307 assumption is made that this pbuf is in one piece (i.e. not chained). These
2361 * api_msg.c: pcb_new sets conn->err if protocol is not implemented
2392 link headers. Now the space is tested and a new pbuf is allocated for the
2427 is only 8 bits wide. This affects the api, as there, the protocol was
2439 * api_msg.c, udp.c: If a udp_pcb has a local_ip set, check if it is the same
2445 with NO_SYS=1". Note that igmp_init is always in tcpip_thread (and not in
2447 added (mac filter is updated only in igmp_init for the moment).
2469 to save a little RAM (next pointer of memp is not used while not in pool).
2480 putting packets on a list that is fed to the stack by calling loopif_poll()
2487 * sockets.h: FD_SETSIZE needs to match number of sockets, which is
2495 timer is reset to fix bug#19434, with help from Oleg Tyshev.
2504 "Constant is long" warnings with 16bit compilers. Contributed by
2509 the mailbox is active". Now, the post is only done during a connect, and do_send,
2530 definition of macros (in errno.h and lwip/arch.h) if LWIP_PROVIDE_ERRNO is
2531 defined. This is the way it should have been already (looking at
2540 to send some garbage. It is not a definitive solution, but the patch does solve
2563 Free tcpip_msg from tcpip_apimsg is not done in tcpip_thread. This give a
2579 this is no worse than before when they would carry on regardless
2618 errno is set inside of get_socket(). Move last sock_set_errno() inside
2664 SYS_LIGHTWEIGHT_PROT=1 & freeing PBUF_RAM when mem_sem is not available)
2673 It is static, and never used in udp.c except udp_init().
2678 initialized before tcpip_thread is running (e.g. before OS is started)
2685 * pbuf.c: Fix BUG#17645 - ensure pbuf payload pointer is not moved
2689 * sockets.c: Fix BUG#19161 - ensure milliseconds timeout is non-zero
2690 when supplied timeout is also non-zero
2701 Note this is a workaround. Currently I have no other options left.
2763 In accept_function() ensure newconn->callback is always initialized.
2784 * tcp_{in,out}.c, tcp_out.c: Removed 'even sndbuf' fix in TCP, which is
2842 * etharp.c: ETHARP_CREATE is renamed to ETHARP_TRY_HARD.
2855 * tcp.c: Increment pcb->snd_buf when ACK is received in SYN_SENT state.
2861 even if one is already pending, if the rcv_wnd is above a threshold
2863 delayed ACK in order to open the window if the stack is only receiving data.
2870 is properly ref counted.
2929 is to prefix the 14-bit Ethernet headers with two padding bytes.
2934 is not queued, and gets lost. Fixed the case where the packets destination
2935 address is already known; we now always queue the packet and perform an ARP
2982 * DEBUGF now is LWIP_DEBUGF
3006 This is the RFC recommended behaviour, but can be overridden in
3037 The application can safely assume the packet is sent, and the RAM
3038 is available to the application directly after calling udp_send()
3049 * IP_ADDR_ANY is no longer a NULL pointer. Instead, it is a pointer
3052 * The packet buffer implementation is changed. The pbuf->ref counter
3210 in 0.4, has now been fixed. Binding to an address that is in use
3248 rapidly allocated and deallocated is now kept in pools. Allocation
3249 and deallocation from those memory pools is very fast. The shared
3250 heap is still present but is used less frequently.