Commit d0df550
committed
bgpd: Free up leaked bpme on release
Direct leak of 160 byte(s) in 1 object(s) allocated from:
0 0x7f5b26cb83b7 in __interceptor_calloc ../../../../src/libsanitizer/asan/asan_malloc_linux.cpp:77
1 0x7f5b2683a34d in qcalloc lib/memory.c:106
2 0x5592a7ec084a in bgp_pbr_match_entry_alloc_intern bgpd/bgp_pbr.c:1097
3 0x7f5b267f9b2f in hash_get lib/hash.c:147
4 0x5592a7ec3654 in bgp_pbr_policyroute_add_to_zebra_unit bgpd/bgp_pbr.c:2467
5 0x5592a7ec643b in bgp_pbr_policyroute_add_to_zebra bgpd/bgp_pbr.c:2608
6 0x5592a7ecd423 in bgp_pbr_handle_entry bgpd/bgp_pbr.c:2813
7 0x5592a7ecd423 in bgp_pbr_update_entry bgpd/bgp_pbr.c:2876
8 0x5592a7fd740a in bgp_zebra_announce_actual bgpd/bgp_zebra.c:1539
9 0x5592a7fdd85d in bgp_handle_route_announcements_to_zebra bgpd/bgp_zebra.c:1810
10 0x7f5b268eecf0 in event_call lib/event.c:2019
11 0x7f5b2681b293 in frr_run lib/libfrr.c:1247
12 0x5592a7dae8b7 in main bgpd/bgp_main.c:557
13 0x7f5b2630c249 in __libc_start_call_main ../sysdeps/nptl/libc_start_call_main.h:58
Turns out that the hash_release of a bpme and not actually
freeing it.... is a recipe to leak memory.
Signed-off-by: Donald Sharp <[email protected]>1 parent 9991f84 commit d0df550
1 file changed
+5
-0
lines changed| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
1683 | 1683 | | |
1684 | 1684 | | |
1685 | 1685 | | |
| 1686 | + | |
| 1687 | + | |
| 1688 | + | |
1686 | 1689 | | |
1687 | 1690 | | |
1688 | 1691 | | |
| |||
1730 | 1733 | | |
1731 | 1734 | | |
1732 | 1735 | | |
| 1736 | + | |
| 1737 | + | |
1733 | 1738 | | |
1734 | 1739 | | |
1735 | 1740 | | |
| |||
0 commit comments