[gmx-users] grompp failure on BG/L

YOLANDA SMALL yas102 at psu.edu
Tue Jul 10 21:58:46 CEST 2007


Hi,

I compiled GROMACS for Blue Gene/L, 'mdrun' seems to be working fine using a
.tpr from a locally built linux version of GROMACS.  However, 'grompp' fails
with a segmentation fault and the following output:

Summary:
program........................../home/ysmall/gromacs-3.3.1/Install_bgl/bin/grompp_smpi
ended with software signal.......0x0000000b (SIGSEGV - segmentation violation)
generated by interrupt...........0x00000002 (data storage interrupt)
while executing instruction at...0x0023d45c
dereferencing memory at..........0x00000000

General Purpose Registers:
r00=0x0023d458 r01=0x1ffb3af0 r02=0xdeadbeef r03=0x00000000
r04=0x1ffb2ae0 r05=0x00000009 r06=0xfefefeff r07=0x7f7f7f7f
r08=0x80808080 r09=0x00000000 r10=0x80808080 r11=0x00000000
r12=0x24000022 r13=0xdeadbeef r14=0x00388190 r15=0x00388188
r16=0x004125dc r17=0x004125ec r18=0x0045fbd0 r19=0x0045f920
r20=0x00000001 r21=0x00000000 r22=0x1ffb7fc0 r23=0x004125d8
r24=0x00410000 r25=0x00410000 r26=0x00413e18 r27=0x00466870
r28=0x003a5c94 r29=0x00000000 r30=0x00000000 r31=0x003a5a90

Special Purpose Registers:
lr=0x0023d458 cr=0x84000024 xer=0x20000002 ctr=0x00000000
msr=0x0002f900 dear=0x00000000 esr=0x00000000 fpscr=0x82002000
dbcr0=0x00000000 dbsr=0x00000000 ccr0=0x00002000
sprg1=0x00080000 sprg2=0x00000000 sprg3=0x1ffb2ae0 sprg6=0x00000000 sprg7=0x00000000

Floating Point Registers:
fpr0=0x00000000 00000000 00000000 00000000
fpr1=0x00000000 00000000 00000000 00000000
fpr2=0x00000000 00000000 00000000 00000000
fpr3=0x00000000 00000000 00000000 00000000
fpr4=0x00000000 00000000 00000000 00000000
fpr5=0x696e5f73 6600326d 00000000 3ff00000
fpr6=0x726f6d70 696e2f67 676c2f62 6c6c5f62
fpr7=0x6e737461 2e312f49 2d332e33 6d616373
fpr8=0x31360042 2d6e7000 74707200 5f31362e
fpr9=0x70695f31 705f736d 726f6d70 6c6c2f67
fpr10=0x00000000 00000000 00000000 00000000
fpr11=0x79736d61 7666732f 6f002f70 726f002d
fpr12=0x00000000 00000000 00000000 00000000
fpr13=0x00000000 5a453d31 495f5349 474c4d50
fpr14=0xffffffff ffffffff ffffffff ffffffff
fpr15=0xffffffff ffffffff ffffffff ffffffff
fpr16=0xffffffff ffffffff ffffffff ffffffff
fpr17=0xffffffff ffffffff ffffffff ffffffff
fpr18=0xffffffff ffffffff ffffffff ffffffff
fpr19=0xffffffff ffffffff ffffffff ffffffff
fpr20=0xffffffff ffffffff ffffffff ffffffff
fpr21=0xffffffff ffffffff ffffffff ffffffff
fpr22=0xffffffff ffffffff ffffffff ffffffff
fpr23=0xffffffff ffffffff ffffffff ffffffff
fpr24=0xffffffff ffffffff ffffffff ffffffff
fpr25=0xffffffff ffffffff ffffffff ffffffff
fpr26=0xffffffff ffffffff ffffffff ffffffff
fpr27=0xffffffff ffffffff ffffffff ffffffff
fpr28=0xffffffff ffffffff ffffffff ffffffff
fpr29=0xffffffff ffffffff ffffffff ffffffff
fpr30=0xffffffff ffffffff ffffffff ffffffff
fpr31=0xffffffff ffffffff 00000000 00000000

Memory:
stack top........................0x1ffc0000
stack frame pointer..............0x1ffb3af0
end of heap......................0x00c86000
start of program.................0x00200000
brk() failed w/ ENOMEM...........0 time(s)

Personality:
XYZT Coordinates.................0, 0, 0, 0
MPI Rank.........................0
mode.............................coprocessor
job id...........................130333

Interrupt History:
number of interrupts.............127
current timebase.................00000003647cfb0e
TB=000000036474070c iar=003441c0 sp=1ffb38d0 value=000000c5 (system call interrupt)
TB=0000000364742b0c iar=00362e74 sp=1ffb3800 value=0000002d (system call interrupt)
TB=00000003647432f4 iar=003441f0 sp=1ffb3970 value=00000003 (system call interrupt)
TB=000000036474ff2c iar=003441f0 sp=1ffb3970 value=00000003 (system call interrupt)
TB=0000000364751ce0 iar=003441e0 sp=1ffb39c0 value=00000006 (system call interrupt)
TB=0000000364754d94 iar=00362e74 sp=1ffb3910 value=0000002d (system call interrupt)
TB=0000000364756d38 iar=00362c04 sp=1ffb26d0 value=0000007a (system call interrupt)
TB=0000000364757c99 iar=0023d45c sp=1ffb3af0 value=00000000 (data storage interrupt)

DCRs:
DCR BGL_TSDCR_RE_SND_XP = 0
DCR BGL_TSDCR_RE_SND_XM = 0
DCR BGL_TSDCR_RE_SND_YP = 0
DCR BGL_TSDCR_RE_SND_YM = 0
DCR BGL_TSDCR_RE_SND_ZP = 0
DCR BGL_TSDCR_RE_SND_ZM = 0

Function Call Chain:
0x0023d45c
0x002392b0
0x002136fc
0x00219440
0x002001e4
End of stack

I recompiled grompp with and without mpi enabled and with a variety of optimization options (-g O2, -g O, etc.) but the jobs still die with segmentation fault errors.  Any suggestions? 

Thanks in advance for your help,
Yolanda




More information about the gromacs.org_gmx-users mailing list