diff options
| author | Pau Espin Pedrol <pespin@sysmocom.de> | 2017-10-16 10:50:43 +0200 |
|---|---|---|
| committer | Pau Espin Pedrol <pespin@sysmocom.de> | 2017-10-17 10:10:44 +0200 |
| commit | 73ab3f070a1fb7fbcf045f528d4109be400ccce6 (patch) | |
| tree | 54f9892b8496264682ad88964b73dc332c59608d | |
| parent | 4fe3ee79a934b7f8d744956feb729723c38fdb7a (diff) | |
bsc_api: Fix NULL secondary_lchan access in handle_ass_fail
Related: OW#3893
Program terminated with signal SIGSEGV, Segmentation fault.
0 gsm_lchan_name (lchan=lchan@entry=0x0) at gsm_data_shared.c:342
(gdb) bt
0 gsm_lchan_name (lchan=lchan@entry=0x0) at gsm_data_shared.c:342
1 0x0805ab80 in lchan_release (lchan=0x0, sacch_deact=sacch_deact@entry=0, mode=mode@entry=RSL_REL_LOCAL_END)
at chan_alloc.c:410
2 0x0805c1dd in handle_ass_fail (msg=0x94142b8, conn=0x9251048) at bsc_api.c:459
3 dispatch_dtap (msg=0x94142b8, link_id=0 '\000', conn=0x9251048) at bsc_api.c:598
4 gsm0408_rcvmsg (msg=msg@entry=0x94142b8, link_id=0 '\000') at bsc_api.c:658
5 0x08058ca2 in abis_rsl_rx_rll (msg=0x94142b8) at abis_rsl.c:1686
6 abis_rsl_rcvmsg (msg=0x94142b8) at abis_rsl.c:2097
7 0xb7e8cf9a in handle_ts1_read (bfd=0x94e8e08) at input/ipaccess.c:271
8 ipaccess_fd_cb (bfd=0x94e8e08, what=1) at input/ipaccess.c:386
9 0xb7ee8434 in osmo_select_main (polling=polling@entry=0) at select.c:158
10 0x0804bd7c in main (argc=6, argv=0xbfc27144) at osmo_bsc_main.c:272
(gdb) print lchan
$2 = (const struct gsm_lchan *) 0x0
Possible scenario in which this crash can appear:
1- gsm0808_assign_req() calls handle_new_assignment() which sends an CHAN
ACTIVATE msg and arms T10 timer.
2- ACTIVATE ACK is received (handle_chan_ack), which calls
gsm48_send_rr_ass_cmd() which sends an ASSIGNMENT CMD, and doesn't
disable/modify T10 timer.
3- T10 timeout is triggered (assignment_t10_timeout()), which sets
conn->secondary_lchan = NULL
4- Immediately after, the ASSIGNMENT FAILURE message (which might have been
already queued) is processed in handle_ass_fail, and then the crash occurs.
This race condition is not an issue for handle_ass_compl() path because there's
this check there which would trigger most probably if secondary_lchan is NULL:
"if (conn->secondary_lchan != msg->lchan)"
Change-Id: I3798b36c628f75d4e8bc7b0996c27d695d53fbb1
| -rw-r--r-- | src/libbsc/bsc_api.c | 6 |
1 files changed, 4 insertions, 2 deletions
diff --git a/src/libbsc/bsc_api.c b/src/libbsc/bsc_api.c index b42c382cb..44c50c13a 100644 --- a/src/libbsc/bsc_api.c +++ b/src/libbsc/bsc_api.c @@ -487,8 +487,10 @@ static void handle_ass_fail(struct gsm_subscriber_connection *conn, /* stop the timer and release it */ osmo_timer_del(&conn->T10); - lchan_release(conn->secondary_lchan, 0, RSL_REL_LOCAL_END); - conn->secondary_lchan = NULL; + if (conn->secondary_lchan) { + lchan_release(conn->secondary_lchan, 0, RSL_REL_LOCAL_END); + conn->secondary_lchan = NULL; + } gh = msgb_l3(msg); if (msgb_l3len(msg) - sizeof(*gh) != 1) { |