bts: Use ms_store when calculating set of target PDCHs for Pkt Paging Request

The ul_tbfs/dl_tbfs lists will become per-trx. Since in this case we
want to operate on the BTS globally, let's iterate over MS objects
instead. This makes more sense too since here we really aim at reaching
a MS (subscriber) instead of specific TBFs. Later on the code can be
optimized easily to schedule a Pkt Paging Request for only 1 of the TBFs
of each MS instad of scheduling it for each TBFs in the MS.

Change-Id: I671e531921bbea2f5cc0f2bfcb8a39ea5c6673b8
Pau Espin 2 years ago
parent bd54205475
commit 9d2fd018ff
  1. 21

@ -373,17 +373,11 @@ int bts_add_paging(struct gprs_rlcmac_bts *bts, const struct paging_req_cs *req,
uint8_t l, trx, ts, any_tbf = 0;
struct gprs_rlcmac_tbf *tbf;
struct llist_item *pos;
struct llist_head *tmp;
const struct osmo_mobile_identity *mi;
uint8_t slot_mask[8];
int8_t first_ts; /* must be signed */
struct llist_head *tbfs_lists[] = {
/* First, build the MI used to page on PDCH from available subscriber info: */
if (req->mi_tmsi_present) {
mi = &req->mi_tmsi;
@ -409,13 +403,18 @@ int bts_add_paging(struct gprs_rlcmac_bts *bts, const struct paging_req_cs *req,
/* We don't know the target MS.
* collect slots to page
* Mark slots for every TBF, but only mark one of it.
* Mark up to one slot attached to each of the TBF of the MS.
* Mark only the first slot found.
* Don't mark, if TBF uses a different slot that is already marked. */
memset(slot_mask, 0, sizeof(slot_mask));
for (l = 0; tbfs_lists[l]; l++) {
llist_for_each_entry(pos, tbfs_lists[l], list) {
tbf = (struct gprs_rlcmac_tbf *)pos->entry;
llist_for_each(tmp, bts_ms_store(bts)->ms_list()) {
ms = llist_entry(tmp, typeof(*ms), list);
struct gprs_rlcmac_tbf *tbfs[] = { ms->ul_tbf, ms->dl_tbf };
for (l = 0; l < ARRAY_SIZE(tbfs); l++) {
tbf = (struct gprs_rlcmac_tbf *)tbfs[l];
if (!tbf)
first_ts = -1;
for (ts = 0; ts < 8; ts++) {
if (tbf->pdch[ts]) {