You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mynewt.apache.org by GitBox <gi...@apache.org> on 2017/10/19 22:25:27 UTC
[GitHub] mkiiskila closed pull request #634: Coap over Lora changes
mkiiskila closed pull request #634: Coap over Lora changes
URL: https://github.com/apache/mynewt-core/pull/634
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull request (from a fork), the diff is supplied
below (as it won't show otherwise due to GitHub magic):
diff --git a/encoding/cborattr/src/cborattr.c b/encoding/cborattr/src/cborattr.c
index 6f6912e2c..096414905 100644
--- a/encoding/cborattr/src/cborattr.c
+++ b/encoding/cborattr/src/cborattr.c
@@ -186,13 +186,13 @@ cbor_internal_read_object(CborValue *root_value,
}
/* contains key value pairs */
- while (cbor_value_is_valid(&cur_value)) {
+ while (cbor_value_is_valid(&cur_value) && !err) {
/* get the attribute */
if (cbor_value_is_text_string(&cur_value)) {
if (cbor_value_calculate_string_length(&cur_value, &len) == 0) {
if (len > MYNEWT_VAL(CBORATTR_MAX_SIZE)) {
err |= CborErrorDataTooLarge;
- goto err_return;
+ break;
}
err |= cbor_value_copy_text_string(&cur_value, attrbuf, &len,
NULL);
@@ -205,7 +205,7 @@ cbor_internal_read_object(CborValue *root_value,
type = cbor_value_get_type(&cur_value);
} else {
err |= CborErrorIllegalType;
- goto err_return;
+ break;
}
} else {
attrbuf[0] = '\0';
@@ -278,9 +278,10 @@ cbor_internal_read_object(CborValue *root_value,
}
cbor_value_advance(&cur_value);
}
-err_return:
- /* that should be it for this container */
- err |= cbor_value_leave_container(root_value, &cur_value);
+ if (!err) {
+ /* that should be it for this container */
+ err |= cbor_value_leave_container(root_value, &cur_value);
+ }
return err;
}
diff --git a/net/oic/src/port/mynewt/lora_adaptor.c b/net/oic/src/port/mynewt/lora_adaptor.c
index 424c687f3..a5f1fd06c 100644
--- a/net/oic/src/port/mynewt/lora_adaptor.c
+++ b/net/oic/src/port/mynewt/lora_adaptor.c
@@ -67,142 +67,182 @@ STATS_NAME_START(oc_lora_stats)
STATS_NAME_END(oc_lora_stats)
static uint8_t oc_lora_stat_reg;
-static STAILQ_HEAD(, os_mbuf_pkthdr) oc_lora_reass_q =
- STAILQ_HEAD_INITIALIZER(oc_lora_reass_q);
+static struct oc_lora_state {
+ struct os_mbuf_pkthdr *rx_pkt;
+ STAILQ_HEAD(, os_mbuf_pkthdr) tx_q;
+ uint8_t tx_frag_num;
+ uint8_t rx_frag_num;
+ uint16_t rx_crc;
+ uint8_t rx_port;
+} oc_lora_state = {
+ .tx_q = STAILQ_HEAD_INITIALIZER(oc_lora_state.tx_q)
+};
/*
* Fragmentation/reassembly.
*/
+#define COAP_LORA_LAST_FRAG 0x80 /* Indicates last frag in a frame. */
+#define COAP_LORA_FRAG_NUM(a) ((a) & ~COAP_LORA_LAST_FRAG)
+
struct coap_lora_frag_start {
uint8_t frag_num; /* 0 */
- uint8_t frag_cnt; /* number of fragments to expect */
uint16_t crc; /* crc over reassembled packet */
-};
+} __attribute__((packed));
struct coap_lora_frag {
uint8_t frag_num; /* which fragment within packet */
};
void
-oc_send_buffer_lora(struct os_mbuf *m)
+oc_send_frag_lora(struct oc_lora_state *os)
{
union {
struct coap_lora_frag_start s;
struct coap_lora_frag f;
} hdr;
struct oc_endpoint_lora *oe;
+ struct os_mbuf_pkthdr *pkt;
+ struct os_mbuf *m;
struct os_mbuf *n;
- int bytes;
- int src_off;
- int off;
+ struct os_mbuf *o;
int mtu;
int blk_len;
- int i;
uint16_t crc;
+ pkt = STAILQ_FIRST(&os->tx_q);
+ m = OS_MBUF_PKTHDR_TO_MBUF(pkt);
oe = (struct oc_endpoint_lora *)OC_MBUF_ENDPOINT(m);
- bytes = OS_MBUF_PKTLEN(m);
+
mtu = lora_app_mtu();
- crc = CRC16_INITIAL_CRC;
- for (n = m; n; n = SLIST_NEXT(n, om_next)) {
- crc = crc16_ccitt(crc, n->om_data, n->om_len);
+ n = os_msys_get_pkthdr(mtu, sizeof(struct lora_pkt_info));
+ if (!n) {
+ goto oom;
}
- i = 0;
- src_off = 0;
- for (off = 0; off < bytes; off += blk_len) {
- n = os_msys_get_pkthdr(mtu, sizeof(struct lora_pkt_info));
- if (!n) {
- goto oom;
+ if (os->tx_frag_num == 0) {
+ crc = CRC16_INITIAL_CRC;
+ for (o = m; o; o = SLIST_NEXT(o, om_next)) {
+ crc = crc16_ccitt(crc, o->om_data, o->om_len);
}
- if (off == 0) {
- hdr.s.frag_num = i++;
- hdr.s.frag_cnt = ((bytes + 2) / (mtu - 1)) + 1;
- hdr.s.crc = crc;
- blk_len = mtu - sizeof(hdr.s);
- if (blk_len > bytes) {
- blk_len = bytes;
- }
- if (os_mbuf_copyinto(n, 0, &hdr.s, sizeof(hdr.s))) {
- goto oom;
- }
- } else {
- hdr.f.frag_num = i++;
- blk_len = mtu - sizeof(hdr.f);
- if (blk_len > bytes - src_off) {
- blk_len = bytes - src_off;
- }
- if (os_mbuf_copyinto(n, 0, &hdr.f, sizeof(hdr.f))) {
- goto oom;
- }
+ hdr.s.frag_num = 0;
+ hdr.s.crc = crc;
+ os->tx_frag_num = 1;
+ blk_len = mtu - sizeof(hdr.s);
+ if (blk_len >= pkt->omp_len) {
+ blk_len = pkt->omp_len;
+ hdr.s.frag_num |= COAP_LORA_LAST_FRAG;
}
- if (os_mbuf_appendfrom(n, m, src_off, blk_len)) {
+ if (os_mbuf_copyinto(n, 0, &hdr.s, sizeof(hdr.s))) {
goto oom;
}
- src_off += blk_len;
- STATS_INC(oc_lora_stats, oframe);
- STATS_INCN(oc_lora_stats, obytes, OS_MBUF_PKTLEN(n));
- if (lora_app_port_send(oe->port, MCPS_CONFIRMED, n)) {
- STATS_INC(oc_lora_stats, oerr);
- goto err;
+ } else {
+ hdr.f.frag_num = os->tx_frag_num++;
+ blk_len = mtu - sizeof(hdr.f);
+ if (blk_len >= pkt->omp_len) {
+ blk_len = pkt->omp_len;
+ hdr.f.frag_num |= COAP_LORA_LAST_FRAG;
+ }
+ if (os_mbuf_copyinto(n, 0, &hdr.f, sizeof(hdr.f))) {
+ goto oom;
}
}
- os_mbuf_free_chain(m);
+ if (os_mbuf_appendfrom(n, m, 0, blk_len)) {
+ goto oom;
+ }
+ os_mbuf_adj(m, blk_len);
+ STATS_INC(oc_lora_stats, oframe);
+ STATS_INCN(oc_lora_stats, obytes, OS_MBUF_PKTLEN(n));
+ if (lora_app_port_send(oe->port, MCPS_CONFIRMED, n)) {
+ STATS_INC(oc_lora_stats, oerr);
+ goto err;
+ }
return;
oom:
STATS_INC(oc_lora_stats, oom);
err:
os_mbuf_free_chain(n);
os_mbuf_free_chain(m);
+ STAILQ_REMOVE_HEAD(&os->tx_q, omp_next);
+ /* XXX unlikely that there's something else queued, but if there is,
+ * we should start tx again */
}
static void
oc_lora_tx_cb(uint8_t port, LoRaMacEventInfoStatus_t status, Mcps_t pkt_type,
struct os_mbuf *m)
{
+ struct oc_lora_state *os = &oc_lora_state;
+ struct os_mbuf_pkthdr *pkt;
+
os_mbuf_free_chain(m);
if (status != LORAMAC_EVENT_INFO_STATUS_OK) {
STATS_INC(oc_lora_stats, oerr);
}
+
+ pkt = STAILQ_FIRST(&os->tx_q);
+ assert(pkt);
+ if (pkt->omp_len == 0) {
+ STAILQ_REMOVE_HEAD(&os->tx_q, omp_next);
+ os->tx_frag_num = 0;
+ os_mbuf_free_chain(OS_MBUF_PKTHDR_TO_MBUF(pkt));
+ }
+ if (!STAILQ_EMPTY(&os->tx_q)) {
+ oc_send_frag_lora(os);
+ }
}
-static struct os_mbuf *
-oc_lora_rx_compact(struct os_mbuf *m)
+void
+oc_send_buffer_lora(struct os_mbuf *m)
+{
+ struct oc_lora_state *os = &oc_lora_state;
+ int in_progress = 0;
+
+ if (!STAILQ_EMPTY(&os->tx_q)) {
+ in_progress = 1;
+ }
+ STAILQ_INSERT_TAIL(&os->tx_q, OS_MBUF_PKTHDR(m), omp_next);
+ if (!in_progress) {
+ oc_send_frag_lora(os);
+ }
+}
+
+static void
+oc_lora_rx_append(struct oc_lora_state *os, struct os_mbuf *m)
{
struct os_mbuf *n;
- n = SLIST_NEXT(m, om_next);
- if (OS_MBUF_TRAILINGSPACE(m) >= n->om_len) {
- memcpy(m->om_data + m->om_len, n->om_data, n->om_len);
- m->om_len += n->om_len;
- SLIST_NEXT(m, om_next) = SLIST_NEXT(n, om_next);
- os_mbuf_free(n);
- return m;
+ /*
+ * Don't try to be too smart about this: if data comes in small
+ * fragments, fragment would likely be in a single incoming mbuf.
+ *
+ * If that fits within the last mbuf in the chain, copy it in.
+ */
+ os->rx_pkt->omp_len += OS_MBUF_PKTLEN(m);
+ n = OS_MBUF_PKTHDR_TO_MBUF(os->rx_pkt);
+ while (SLIST_NEXT(n, om_next)) {
+ n = SLIST_NEXT(n, om_next);
+ }
+ if (OS_MBUF_TRAILINGSPACE(n) >= m->om_len) {
+ memcpy(n->om_data + n->om_len, m->om_data, m->om_len);
+ n->om_len += m->om_len;
+ SLIST_NEXT(n, om_next) = SLIST_NEXT(m, om_next);
+ os_mbuf_free(m);
} else {
- return n;
+ SLIST_NEXT(n, om_next) = m;
}
}
static void
-oc_lora_deliver(struct os_mbuf_pkthdr *mp, uint16_t port)
+oc_lora_deliver(struct oc_lora_state *os)
{
struct oc_endpoint_lora *oe;
struct os_mbuf *m;
struct os_mbuf *n;
- struct os_mbuf *o;
- struct os_mbuf_pkthdr *np;
uint16_t crc;
- uint16_t pktcrc;
- m = OS_MBUF_PKTHDR_TO_MBUF(mp);
+ m = OS_MBUF_PKTHDR_TO_MBUF(os->rx_pkt);
- os_mbuf_copydata(m, (int)(&((struct coap_lora_frag_start *)0)->crc),
- sizeof(crc), &pktcrc);
-
- /*
- * Trim frag header out.
- */
- os_mbuf_adj(m, sizeof(struct coap_lora_frag_start));
+ os->rx_pkt = NULL;
/*
* add oc_endpoint_lora in the front.
@@ -221,33 +261,16 @@ oc_lora_deliver(struct os_mbuf_pkthdr *mp, uint16_t port)
}
oe = (struct oc_endpoint_lora *)OC_MBUF_ENDPOINT(m);
oe->flags = LORA;
- oe->port = port;
+ oe->port = os->rx_port;
/*
- * CRC first fragment.
+ * Check CRC.
*/
- crc = crc16_ccitt(CRC16_INITIAL_CRC, m->om_data, m->om_len);
- n = m;
- while ((o = SLIST_NEXT(n, om_next))) {
- crc = crc16_ccitt(crc, o->om_data, o->om_len);
- n = oc_lora_rx_compact(n);
- }
-
- while ((np = STAILQ_NEXT(mp, omp_next))) {
- /*
- * XXX coalesce short fragments
- */
- STAILQ_NEXT(mp, omp_next) = STAILQ_NEXT(np, omp_next);
- os_mbuf_adj(OS_MBUF_PKTHDR_TO_MBUF(np), sizeof(struct coap_lora_frag));
- SLIST_NEXT(n, om_next) = OS_MBUF_PKTHDR_TO_MBUF(np);
- mp->omp_len += np->omp_len;
- while ((o = SLIST_NEXT(n, om_next))) {
- crc = crc16_ccitt(crc, o->om_data, o->om_len);
- n = oc_lora_rx_compact(n);
- }
+ crc = CRC16_INITIAL_CRC;
+ for (n = m; n; n = SLIST_NEXT(n, om_next)) {
+ crc = crc16_ccitt(crc, n->om_data, n->om_len);
}
-
- if (crc != pktcrc) {
+ if (crc != os->rx_crc) {
STATS_INC(oc_lora_stats, icsum);
os_mbuf_free_chain(m);
} else {
@@ -255,37 +278,21 @@ oc_lora_deliver(struct os_mbuf_pkthdr *mp, uint16_t port)
}
}
-static uint8_t
-oc_lora_frag_num(struct os_mbuf *m)
-{
- struct coap_lora_frag *cf;
-
- cf = (struct coap_lora_frag *)m->om_data;
- return cf->frag_num;
-}
-
-static uint8_t
-oc_lora_frag_cnt(struct os_mbuf *m)
-{
- struct coap_lora_frag_start *cfs;
-
- cfs = (struct coap_lora_frag_start *)m->om_data;
- return cfs->frag_cnt;
-}
-
static void
oc_lora_rx_reass(struct os_mbuf *m, uint16_t port)
{
+ struct oc_lora_state *os = &oc_lora_state;
struct coap_lora_frag_start *cfs;
struct coap_lora_frag cf;
+ uint8_t frag_num;
struct os_mbuf_pkthdr *pkt;
- struct os_mbuf_pkthdr *pkt2;
if (os_mbuf_copydata(m, 0, sizeof(cf), &cf)) {
STATS_INC(oc_lora_stats, ishort);
goto err;
}
- if (cf.frag_num == 0) {
+ frag_num = COAP_LORA_FRAG_NUM(cf.frag_num);
+ if (frag_num == 0) {
m = os_mbuf_pullup(m, sizeof(*cfs));
} else {
m = os_mbuf_pullup(m, sizeof(cf));
@@ -297,35 +304,40 @@ oc_lora_rx_reass(struct os_mbuf *m, uint16_t port)
pkt = OS_MBUF_PKTHDR(m);
- if (STAILQ_EMPTY(&oc_lora_reass_q)) {
- if (cf.frag_num != 0) {
+ if (os->rx_pkt == NULL) {
+ /*
+ * If no frame being reassembled, then it must have frag num 0,
+ * and start header.
+ */
+ if (frag_num != 0) {
STATS_INC(oc_lora_stats, ioof);
goto err;
}
- if (oc_lora_frag_cnt(m) == 1) {
- oc_lora_deliver(pkt, port);
- } else {
- STAILQ_INSERT_TAIL(&oc_lora_reass_q, pkt, omp_next);
- }
+ os_mbuf_copydata(m, (int)(&((struct coap_lora_frag_start *)0)->crc),
+ sizeof(os->rx_crc), &os->rx_crc);
+ os_mbuf_adj(m, sizeof(struct coap_lora_frag_start));
+ os->rx_frag_num = 0;
+ os->rx_pkt = pkt;
+ os->rx_port = port;
} else {
- pkt2 = STAILQ_LAST(&oc_lora_reass_q, os_mbuf_pkthdr, omp_next);
- if (oc_lora_frag_num(OS_MBUF_PKTHDR_TO_MBUF(pkt2)) + 1 !=
- cf.frag_num) {
- while ((pkt2 = STAILQ_FIRST(&oc_lora_reass_q))) {
- STAILQ_REMOVE_HEAD(&oc_lora_reass_q, omp_next);
- STATS_INC(oc_lora_stats, ioof);
- os_mbuf_free_chain(OS_MBUF_PKTHDR_TO_MBUF(pkt2));
- }
- STAILQ_INSERT_TAIL(&oc_lora_reass_q, pkt, omp_next);
- } else {
- STAILQ_INSERT_TAIL(&oc_lora_reass_q, pkt, omp_next);
- pkt2 = STAILQ_FIRST(&oc_lora_reass_q);
- if (cf.frag_num + 1 ==
- oc_lora_frag_cnt(OS_MBUF_PKTHDR_TO_MBUF(pkt2))) {
- STAILQ_INIT(&oc_lora_reass_q);
- oc_lora_deliver(pkt2, port);
- }
+ /*
+ * Subsequent fragments must come in order.
+ */
+ if (frag_num != os->rx_frag_num + 1 || os->rx_port != port) {
+ os_mbuf_free_chain(OS_MBUF_PKTHDR_TO_MBUF(os->rx_pkt));
+ os->rx_pkt = NULL;
+ STATS_INC(oc_lora_stats, ioof);
+ goto err;
}
+ os->rx_frag_num++;
+ os_mbuf_adj(m, sizeof(struct coap_lora_frag));
+ oc_lora_rx_append(os, m);
+ }
+ if (cf.frag_num & COAP_LORA_LAST_FRAG) {
+ /*
+ * If last fragment, deliver the frame to stack.
+ */
+ oc_lora_deliver(os);
}
return;
err:
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
With regards,
Apache Git Services