| /nrf52832-nimble/packages/NimBLE-latest/nimble/controller/src/ |
| H A D | ble_ll_sched.c | 55 * 1) Add some accounting to the schedule code to see how late we are
58 * 2) Need to determine how we really want to handle the case when we execute
59 * a schedule item but there is a current event. We could:
106 * item. We only care about connection schedule items
183 /* Better be past current time or we just leave */ in ble_ll_sched_conn_reschedule()
188 /* We have to find a place for this schedule */ in ble_ll_sched_conn_reschedule()
196 /* Stop timer since we will add an element */ in ble_ll_sched_conn_reschedule()
205 /* Do nothing, we start_mark overlap below */ in ble_ll_sched_conn_reschedule()
208 /* Only insert if this element is older than all that we in ble_ll_sched_conn_reschedule()
341 * makes us off ~13 usecs. Since we dont want to actually calculate the in ble_ll_sched_master_new()
[all …]
|
| H A D | ble_ll_conn.c | 50 * 1) I think if we are initiating and we already have a connection with
51 * a device that we will still try and connect to it. Fix this.
53 * i) When a connection create is issued, if we already are connected
55 * ii) If we receive an advertisement while initiating and want to send
56 * a connect request to the device, make sure we dont have it.
62 * 2) Make sure we check incoming data packets for size and all that. You
65 * 3) Make sure we are setting the schedule end time properly for both slave
66 * and master. We should just set this to the end of the connection event.
67 * We might want to guarantee a IFS time as well since the next event needs
81 * 7) How does peer address get set if we are using whitelist? Look at filter
[all …]
|
| H A D | ble_ll_ctrl.c | 50 * 2) Should we create pool of control pdu's?. Dont need more
52 * if we have lots of connections.
58 * 5) We are supposed to remember when we do the data length update proc if
59 * the device sent us an unknown rsp. We should not send it another len req.
61 * 8) How to count control pdus sent. DO we count enqueued + sent, or only
62 * sent (actually attempted to tx). Do we count failures? How?
218 * NOTE: we dont have to reset the callout since this packet will have had a
231 * Called when we receive either a connection parameter request or response.
279 * or latency. If not, this may just be an anchor point change and we do in ble_ll_ctrl_conn_param_pdu_proc()
281 * XXX: what if we dont like the parameters? When do we check that out? in ble_ll_ctrl_conn_param_pdu_proc()
[all …]
|
| H A D | ble_ll_conn_hci.c | 38 * Used to limit the rate at which we send the number of completed packets
39 * event to the host. This is the os time at which we can send an event.
167 /* If it was direct advertising we were replying to and we used in ble_ll_conn_comp_event_send()
168 * identity address there (which might be just fine), we should in ble_ll_conn_comp_event_send()
169 * we should take it into account here in this event. in ble_ll_conn_comp_event_send()
184 /* We need to adjust peer type if device connected using RPA in ble_ll_conn_comp_event_send()
246 * At some periodic rate, make sure we go through all active connections in ble_ll_conn_num_comp_pkts_event_send()
247 * and send the number of completed packet events. We do this mainly in ble_ll_conn_num_comp_pkts_event_send()
248 * because the spec says we must update the host even though no packets in ble_ll_conn_num_comp_pkts_event_send()
256 * We do this to increase throughput but we dont want to search the in ble_ll_conn_num_comp_pkts_event_send()
[all …]
|
| H A D | ble_ll_scan.c | 46 * 1) I think I can guarantee that we dont process things out of order if
100 /* We support one report per event for now */
139 /* Contains list of advertisers that we have heard scan responses from */
285 /* We'll use the same timeout as for RPA rotation */ in ble_ll_scan_refresh_nrpa()
347 * If advertising device is on our resolving list, we use RPA generated in ble_ll_scan_req_pdu_make()
349 * we use NRPA as ScanA as allowed by spec to prevent our device from in ble_ll_scan_req_pdu_make()
387 /* Do we have an address match? Must match address type */ in ble_ll_scan_find_dup_adv()
433 /* We support only one report per event now */ in ble_ll_scan_init_ext_adv_report()
507 * (truncated was sent to the host) and we just need to drop last reference. in ble_ll_scan_end_adv_evt()
508 * Otherwise we should try to send truncated event to the host. in ble_ll_scan_end_adv_evt()
[all …]
|
| H A D | ble_ll_adv.c | 49 * 4) How to determine the advertising interval we will actually use. As of
50 * now, we set it to max.
75 * This is the address that we send in directed advertisements (the
76 * INITA field). If we are using Privacy this is a RPA that we need to
77 * generate. We reserve space in the advsm to save time when creating
232 * resolving list. That is why we need to check if the generated address is
262 /* Mark RPA as timed out so we get a new RPA */ in ble_ll_adv_rpa_timeout()
270 * Calculate the first channel that we should advertise upon when we start
295 * Calculate the final channel that we should advertise upon when we start
504 /* We do not create scannable PDUs here - this is handled separately */ in ble_ll_adv_aux_pdu_make()
[all …]
|
| /nrf52832-nimble/rt-thread/components/net/lwip-2.0.2/src/netif/ppp/ |
| H A D | PPPD_FOLLOWUP | 5 is getting low on their Git repository, meaning that we can follow what
8 So, here is the pppd follow up, so that we don't get away too far from pppd.
13 This has nothing to do with pppd, but we merged some good patch from
16 - LCP adaptive echo, so that we don't send LCP echo request if we
64 We can't fork processes in embedded, therefore all the pppd process run
65 feature is disabled in the port, so we don't care about the new
73 Again, we are not running as a heavy process, so all exit() or _exit() calls
81 We are not parsing config file, all the filesystem I/O stuff were disabled
88 Only concern changes on RP-PPPoE plugin, which we don't use.
106 Patch for sys-linux.c, which we don't use.
[all …]
|
| H A D | vj.c | 203 /* TCP stack requires that we don't change the packet payload, therefore we copy in vj_compress_tcp()
219 * Packet is compressible -- we're going to send either a in vj_compress_tcp()
220 * COMPRESSED_TCP or UNCOMPRESSED_TCP packet. Either way we need in vj_compress_tcp()
223 * again & we don't have to do any reordering if it's used. in vj_compress_tcp()
237 * we want is almost always near the front, we locate in vj_compress_tcp()
238 * states via linear search. If we don't find a state in vj_compress_tcp()
257 * connection number we're using for this conversation. in vj_compress_tcp()
259 * state points to the newest and we only need to set in vj_compress_tcp()
283 * Make sure that only what we expect to change changed. The first in vj_compress_tcp()
290 * different between the previous & current datagram, we send the in vj_compress_tcp()
[all …]
|
| H A D | lcp.c | 66 * When the link comes up we want to be able to wait for a short while,
68 * configure-requests. We do this by delaying the fsm_lowerup call.
225 static void lcp_up(fsm *f); /* We're UP */
226 static void lcp_down(fsm *f); /* We're DOWN */
227 static void lcp_starting (fsm *); /* We need lower layer up */
228 static void lcp_finished (fsm *); /* We need lower layer down */
254 lcp_starting, /* Called when we want the lower layer up */
255 lcp_finished, /* Called when we want the lower layer down */
445 * because we are in passive/silent mode or because we have in lcp_close()
463 * if we are going to ask for A/C and protocol compression. in lcp_lowerup()
[all …]
|
| /nrf52832-nimble/rt-thread/components/net/lwip-2.1.0/src/netif/ppp/ |
| H A D | PPPD_FOLLOWUP | 5 is getting low on their Git repository, meaning that we can follow what
8 So, here is the pppd follow up, so that we don't get away too far from pppd.
13 This has nothing to do with pppd, but we merged some good patch from
16 - LCP adaptive echo, so that we don't send LCP echo request if we
64 We can't fork processes in embedded, therefore all the pppd process run
65 feature is disabled in the port, so we don't care about the new
73 Again, we are not running as a heavy process, so all exit() or _exit() calls
81 We are not parsing config file, all the filesystem I/O stuff were disabled
88 Only concern changes on RP-PPPoE plugin, which we don't use.
106 Patch for sys-linux.c, which we don't use.
[all …]
|
| H A D | vj.c | 203 /* TCP stack requires that we don't change the packet payload, therefore we copy in vj_compress_tcp()
214 * Packet is compressible -- we're going to send either a in vj_compress_tcp()
215 * COMPRESSED_TCP or UNCOMPRESSED_TCP packet. Either way we need in vj_compress_tcp()
218 * again & we don't have to do any reordering if it's used. in vj_compress_tcp()
232 * we want is almost always near the front, we locate in vj_compress_tcp()
233 * states via linear search. If we don't find a state in vj_compress_tcp()
252 * connection number we're using for this conversation. in vj_compress_tcp()
254 * state points to the newest and we only need to set in vj_compress_tcp()
278 * Make sure that only what we expect to change changed. The first in vj_compress_tcp()
285 * different between the previous & current datagram, we send the in vj_compress_tcp()
[all …]
|
| /nrf52832-nimble/rt-thread/components/net/uip/uip/ |
| H A D | uip.c | 71 * The principle is that we have a small buffer, called the uip_buf,
566 /* If ip_reasstmr is zero, no packet is present in the buffer, so we in uip_reass()
578 in the reasembly buffer. If so, we proceed with copying the in uip_reass()
591 reassembly buffer, we discard the entire packet. */ in uip_reass()
606 /* If the two endpoints are in the same byte, we only update in uip_reass()
613 /* If the two endpoints are in different bytes, we update the in uip_reass()
625 /* If this fragment has the More Fragments flag set to zero, we in uip_reass()
626 know that this is the last fragment, so we can calculate the in uip_reass()
627 size of the entire packet. We also set the in uip_reass()
628 IP_REASS_FLAG_LASTFRAG flag to indicate that we have received in uip_reass()
[all …]
|
| /nrf52832-nimble/packages/NimBLE-latest/nimble/drivers/nrf52/src/ |
| H A D | ble_phy.c | 59 * and 16ms for a 30ppm crystal! We need to limit PDU size based on
71 * We configure the nrf with a 1 byte S0 field, 8 bit length field, and
127 /* XXX: if 27 byte packets desired we can make this smaller */
209 * pending bit was TRUE for the radio interrupt (as expected) as we never
222 * Not sure what the end result of all this is. We will clear the pending
223 * bit in the NVIC just to be sure when we disable the PHY.
231 * I figure this out I am just going to allocate 67 words so we have enough
471 * Called when we want to wait if the radio is in either the rx or tx
472 * disable states. We want to wait until that state is over before doing
506 * We need to adjust start time to include radio ramp-up and TX pipeline in ble_phy_set_start_time()
[all …]
|
| /nrf52832-nimble/rt-thread/components/net/lwip-2.1.0/src/core/ |
| H A D | tcp_in.c | 246 /* Demultiplex an incoming segment. First, we check if it is destined
267 lookups will be faster (we exploit locality in TCP segment
284 /* If it did not go to an active connection, we check the connections
299 /* We don't really care enough to move this PCB to the front
300 of the list since we are not very likely to receive that
315 /* Finally, if we still did not get a match, we check all PCBs that
362 lookups will be faster (we exploit locality in TCP segment
428 /* this is a zero-window probe, we respond to it with current RCV.NXT
440 and that the pcb has been freed. If so, we don't do anything. */
444 end. We then call the error callback to inform the
[all …]
|
| /nrf52832-nimble/packages/NimBLE-latest/nimble/drivers/nrf51/src/ |
| H A D | ble_phy.c | 42 * and 16ms for a 30ppm crystal! We need to limit PDU size based on
54 * We configure the nrf with a 1 byte S0 field, 8 bit length field, and
96 /* XXX: if 27 byte packets desired we can make this smaller */
160 * pending bit was TRUE for the radio interrupt (as expected) as we never
173 * Not sure what the end result of all this is. We will clear the pending
174 * bit in the NVIC just to be sure when we disable the PHY.
273 * Called when we want to wait if the radio is in either the rx or tx
274 * disable states. We want to wait until that state is over before doing
284 * we use RX symbols only. in nrf_wait_disabled()
323 * With the 32.768 kHz crystal, we may need to adjust the RTC compare in ble_phy_set_start_time()
[all …]
|
| /nrf52832-nimble/rt-thread/components/net/lwip-2.1.0/src/apps/http/ |
| H A D | httpd.c | 542 /* We cannot send more data than space available in the send buffer. */ in http_write()
696 /* If we have no parameters at all, return immediately. */ in extract_uri_parameters()
721 /* We didn't find a new parameter so find the end of the URI and in extract_uri_parameters()
728 /* Revert to NULL so that we exit the loop as expected. */ in extract_uri_parameters()
789 /* Find this tag in the list we have been provided. */
816 /* If we drop out, we were asked to serve a page which contains tags that
817 * we don't have a handler for. Merely echo back the tags with an error
849 /* In all cases, the second header we send is the server identification
855 /* Is this a normal file or the special case we use to send back the
873 /* We are dealing with a particular filename. Look for one other
[all …]
|
| /nrf52832-nimble/rt-thread/components/net/lwip-2.1.0/src/include/lwip/ |
| H A D | ip6_frag.h | 63 * extension header. This means that for each first fragment packet, we need to
64 * 1) make a copy of some IPv6 header fields (src+dest) that we need later on,
65 * just in case we do overwrite part of the IPv6 header, and 2) make a copy of
66 * the header data that we overwrote, so that we can restore it before either
68 * if this setting is disabled, but if it is enabled, we need to save a bit
69 * more data (up to the size of a pointer) because we overwrite more. */
76 * we copy the fields that we need from the IP header for as long as the helper
78 * those fields in the IP header each time we need to perform checks on them. */
97 /* This buffer (for the part of the original header that we overwrite) will
98 * be slightly oversized, but we cannot compute the exact size from here. */
[all …]
|
| /nrf52832-nimble/rt-thread/components/dfs/filesystems/jffs2/src/ |
| H A D | gc.c | 49 /* Pick an eraseblock to garbage collect next. This is where we'll in jffs2_find_gc_block()
51 /* We possibly want to favour the dirtier blocks more when the in jffs2_find_gc_block()
105 /* Have we accidentally picked a clean block with wasted space ? */ in jffs2_find_gc_block()
137 /* We can't start doing GC yet. We haven't finished checking in jffs2_garbage_collect_pass()
180 /* We need to wait for it to finish, lest we move on in jffs2_garbage_collect_pass()
181 and trigger the BUG() above while we haven't yet in jffs2_garbage_collect_pass()
208 /* First, work out which block we're garbage-collecting */ in jffs2_garbage_collect_pass()
252 we don't grok that has JFFS2_NODETYPE_RWCOMPAT_COPY, we should do so */ in jffs2_garbage_collect_pass()
261 /* We need to hold the inocache. Either the erase_completion_lock or in jffs2_garbage_collect_pass()
262 the inocache_lock are sufficient; we trade down since the inocache_lock in jffs2_garbage_collect_pass()
[all …]
|
| /nrf52832-nimble/rt-thread/components/net/lwip-2.0.2/src/apps/httpd/ |
| H A D | httpd.c | 528 /* We cannot send more data than space available in the send buffer. */ in http_write()
682 /* If we have no parameters at all, return immediately. */ in extract_uri_parameters()
707 /* We didn't find a new parameter so find the end of the URI and in extract_uri_parameters()
714 /* Revert to NULL so that we exit the loop as expected. */ in extract_uri_parameters()
775 /* Find this tag in the list we have been provided. */
802 /* If we drop out, we were asked to serve a page which contains tags that
803 * we don't have a handler for. Merely echo back the tags with an error
836 /* In all cases, the second header we send is the server identification
842 /* Is this a normal file or the special case we use to send back the
860 /* We are dealing with a particular filename. Look for one other
[all …]
|
| /nrf52832-nimble/rt-thread/components/net/lwip-2.0.2/src/core/ |
| H A D | tcp_in.c | 219 /* Demultiplex an incoming segment. First, we check if it is destined
232 lookups will be faster (we exploit locality in TCP segment
249 /* If it did not go to an active connection, we check the connections
257 /* We don't really care enough to move this PCB to the front
258 of the list since we are not very likely to receive that
267 /* Finally, if we still did not get a match, we check all PCBs that
307 lookups will be faster (we exploit locality in TCP segment
360 /* this is a zero-window probe, we respond to it with current RCV.NXT
372 and that the pcb has been freed. If so, we don't do anything. */
376 end. We then call the error callback to inform the
[all …]
|
| /nrf52832-nimble/rt-thread/components/net/lwip-1.4.1/src/core/ |
| H A D | tcp_in.c | 165 /* Demultiplex an incoming segment. First, we check if it is destined in tcp_input()
180 lookups will be faster (we exploit locality in TCP segment in tcp_input()
195 /* If it did not go to an active connection, we check the connections in tcp_input()
203 /* We don't really care enough to move this PCB to the front in tcp_input()
204 of the list since we are not very likely to receive that in tcp_input()
213 /* Finally, if we still did not get a match, we check all PCBs that in tcp_input()
247 lookups will be faster (we exploit locality in TCP segment in tcp_input()
306 and that the pcb has been freed. If so, we don't do anything. */ in tcp_input()
310 end. We then call the error callback to inform the in tcp_input()
311 application that the connection is dead before we in tcp_input()
[all …]
|
| /nrf52832-nimble/rt-thread/components/net/lwip-1.4.1/src/netif/ppp/ |
| H A D | lcp.c | 99 lcp_options lcp_wantoptions[NUM_PPP]; /* Options that we want to request */
101 lcp_options lcp_allowoptions[NUM_PPP]; /* Options we allow peer to request */
102 lcp_options lcp_hisoptions[NUM_PPP]; /* Options that we ack'd */
109 /* @todo: do we really need such a large buffer? The typical 1500 bytes seem too much. */
110 static u_char nak_buffer[PPP_MRU]; /* where we construct a nak packet */
122 static void lcp_up (fsm*); /* We're UP */
123 static void lcp_down (fsm*); /* We're DOWN */
124 static void lcp_starting (fsm*); /* We need lower layer up */
125 static void lcp_finished (fsm*); /* We need lower layer down */
151 lcp_starting, /* Called when we want the lower layer up */
[all …]
|
| H A D | vj.c | 171 * Packet is compressible -- we're going to send either a in vj_compress_tcp()
172 * COMPRESSED_TCP or UNCOMPRESSED_TCP packet. Either way we need in vj_compress_tcp()
175 * again & we don't have to do any reordering if it's used. in vj_compress_tcp()
189 * we want is almost always near the front, we locate in vj_compress_tcp()
190 * states via linear search. If we don't find a state in vj_compress_tcp()
209 * connection number we're using for this conversation. in vj_compress_tcp()
211 * state points to the newest and we only need to set in vj_compress_tcp()
248 * Make sure that only what we expect to change changed. The first in vj_compress_tcp()
255 * different between the previous & current datagram, we send the in vj_compress_tcp()
280 * doesn't prohibit the change so we have to deal in vj_compress_tcp()
[all …]
|
| /nrf52832-nimble/rt-thread/components/net/uip/apps/hello-world/ |
| H A D | hello-world.c | 20 * We define the application state (struct hello_world_state) in the
21 * hello-world.h file, so we need to include it here. We also include
23 * <string.h>, since we use the memcpy() function in the code.
37 * The initialization function. We must explicitly call this function
44 /* We start to listen for connections on TCP port 1000. */ in hello_world_init()
49 * In hello-world.h we have defined the UIP_APPCALL macro to
60 * the application state of the connection. We make a pointer to in hello_world_appcall()
66 * If a new connection was just established, we should initialize in hello_world_appcall()
75 * Finally, we run the protosocket function that actually handles in hello_world_appcall()
76 * the communication. We pass it a pointer to the application state in hello_world_appcall()
|
| /nrf52832-nimble/rt-thread/components/net/lwip-2.1.0/src/core/ipv6/ |
| H A D | ip6_frag.c | 74 /* The number of bytes we need to "borrow" from (i.e., overwrite in) the header
163 /* Restore the part that we've overwritten with our helper structure, or we in ip6_reass_free_complete_datagram()
166 /* Then, move back to the original ipv6 header (we are now pointing to Fragment header). in ip6_reass_free_complete_datagram()
167 This cannot fail since we already checked when receiving this fragment. */ in ip6_reass_free_complete_datagram()
172 /* Reconstruct the zoned source and destination addresses, so that we do in ip6_reass_free_complete_datagram()
320 in the reassembly buffer. If so, we proceed with copying the in ip6_reass()
362 * Eventually, we will replace it when we get the first fragment in ip6_reass()
373 * stored zones end up resolving to a different interface. In that case, we in ip6_reass()
387 /* Check if we are allowed to enqueue more datagrams. */ in ip6_reass()
412 This cannot fail since we already checked when receiving this fragment. */ in ip6_reass()
[all …]
|