While looking into another problem I ran into an issue which made ob1 segfault
on me. Using gm, and running the test test_dan1 in the onesided test suite,
if I limit the gm freelist by too much, I get a segfault. That is,
mpirun -np 2 -mca btl gm,self -mca btl_gm_free_list_max 1024 test_dan1
works fine, but
mpirun -np 2 -mca btl gm,self -mca btl_gm_free_list_max 512 test_dan1
segfaults. Here is the relevant output from gdb:
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 1077541088 (LWP 15600)]
0x404d81c1 in mca_pml_ob1_send_fin (proc=0x9bd9490, bml_btl=0xd323580,
hdr_des=0x9e54e78, order=255 '�', status=1) at pml_ob1.c:267
267 MCA_PML_OB1_DES_ALLOC(bml_btl, fin, order,
sizeof(mca_pml_ob1_fin_hdr_t));
(gdb) bt
#0 0x404d81c1 in mca_pml_ob1_send_fin (proc=0x9bd9490, bml_btl=0xd323580,
hdr_des=0x9e54e78, order=255 '�', status=1) at pml_ob1.c:267
#1 0x404eef7a in mca_pml_ob1_send_request_put_frag (frag=0xa711f00)
at pml_ob1_sendreq.c:1141
#2 0x404d986e in mca_pml_ob1_process_pending_rdma () at pml_ob1.c:387
#3 0x404eed57 in mca_pml_ob1_put_completion (btl=0x9c37e38, ep=0x9c42c78,
des=0xb62ad00, status=0) at pml_ob1_sendreq.c:1108
#4 0x404ff520 in mca_btl_gm_put_callback (port=0x9bec5e0, context=0xb62ad00,
status=GM_SUCCESS) at btl_gm.c:682
#5 0x40512c4f in gm_handle_sent_tokens (p=0x9bec5e0, e=0x406189c0)
at ./libgm/gm_handle_sent_tokens.c:82
#6 0x40517c73 in _gm_unknown (p=0x9bec5e0, e=0x406189c0)
at ./libgm/gm_unknown.c:222
#7 0x405180fc in gm_unknown (p=0x9bec5e0, e=0x406189c0)
at ./libgm/gm_unknown.c:300
#8 0x40502708 in mca_btl_gm_component_progress () at btl_gm_component.c:649
#9 0x404f6fd6 in mca_bml_r2_progress () at bml_r2.c:110
#10 0x401a51d3 in opal_progress () at runtime/opal_progress.c:201
#11 0x405cf864 in opal_condition_wait (c=0x9e564b8, m=0x9e56478)
at ../../../../opal/threads/condition.h:98
#12 0x405cf68e in ompi_osc_pt2pt_module_fence (assert=0, win=0x9e55ec8)
at osc_pt2pt_sync.c:142
#13 0x400b6ebb in PMPI_Win_fence (assert=0, win=0x9e55ec8) at pwin_fence.c:57
#14 0x0804a2f3 in test_bandwidth1 (nbufsize=1050000, min_iterations=10,
max_iterations=1000, verbose=0) at test_dan1.c:282
#15 0x0804b06f in get_bandwidth (argc=0, argv=0x0) at test_dan1.c:686
#16 0x080512f5 in test_dan1 () at test_dan1.c:3555
#17 0x08051573 in main (argc=1, argv=0xbfeba9f4) at test_dan1.c:3639
(gdb)
This is using the trunk. Any ideas?
Thanks,
Tim
