*** glibc detected ***

[DTokarev]

Hi,
After last update (rev579M) I received error message:

DEF_HIS - creating history file: ocean_his.nc
*** glibc detected *** ./oceanO: free(): invalid pointer: 0x0000000002385120 ***
======= Backtrace: =========
/lib64/libc.so.6[0x3a21476d63]
./oceanO[0x55ec63]
./oceanO[0x465a0c]
./oceanO[0x448c24]
./oceanO[0x40223a]
./oceanO[0x4020ed]
/lib64/libc.so.6(__libc_start_main+0xfd)[0x3a2141ee5d]
./oceanO[0x401f49]
======= Memory map: ========
00400000-0073f000 r-xp 00000000 08:09 3547429 /home/tdv1/oceanO
0093f000-00941000 rw-p 0033f000 08:09 3547429 /home/tdv1/oceanO
00941000-009e9000 rw-p 00000000 00:00 0
01ec8000-023b7000 rw-p 00000000 00:00 0 [heap]
3a21000000-3a2101f000 r-xp 00000000 08:09 3682518 /lib64/ld-2.13.so
3a2121e000-3a2121f000 r--p 0001e000 08:09 3682518 /lib64/ld-2.13.so
3a2121f000-3a21220000 rw-p 0001f000 08:09 3682518 /lib64/ld-2.13.so
3a21220000-3a21221000 rw-p 00000000 00:00 0
3a21400000-3a21591000 r-xp 00000000 08:09 3682519 /lib64/libc-2.13.so
3a21591000-3a21791000 ---p 00191000 08:09 3682519 /lib64/libc-2.13.so
3a21791000-3a21795000 r--p 00191000 08:09 3682519 /lib64/libc-2.13.so
3a21795000-3a21796000 rw-p 00195000 08:09 3682519 /lib64/libc-2.13.so
3a21796000-3a2179c000 rw-p 00000000 00:00 0
3a21800000-3a21884000 r-xp 00000000 08:09 3682520 /lib64/libm-2.13.so
3a21884000-3a21a83000 ---p 00084000 08:09 3682520 /lib64/libm-2.13.so
3a21a83000-3a21a84000 r--p 00083000 08:09 3682520 /lib64/libm-2.13.so
3a21a84000-3a21a85000 rw-p 00084000 08:09 3682520 /lib64/libm-2.13.so
3a21c00000-3a21c17000 r-xp 00000000 08:09 3682524 /lib64/libpthread-2.13.so
3a21c17000-3a21e16000 ---p 00017000 08:09 3682524 /lib64/libpthread-2.13.so
3a21e16000-3a21e17000 r--p 00016000 08:09 3682524 /lib64/libpthread-2.13.so
3a21e17000-3a21e18000 rw-p 00017000 08:09 3682524 /lib64/libpthread-2.13.so
3a21e18000-3a21e1c000 rw-p 00000000 00:00 0
3a22400000-3a22415000 r-xp 00000000 08:09 3682540 /lib64/libgcc_s-4.5.1-20100924.so.1
3a22415000-3a22614000 ---p 00015000 08:09 3682540 /lib64/libgcc_s-4.5.1-20100924.so.1
3a22614000-3a22615000 rw-p 00014000 08:09 3682540 /lib64/libgcc_s-4.5.1-20100924.so.1
3a22c00000-3a22c07000 r-xp 00000000 08:09 3682531 /lib64/librt-2.13.so
3a22c07000-3a22e06000 ---p 00007000 08:09 3682531 /lib64/librt-2.13.so
3a22e06000-3a22e07000 r--p 00006000 08:09 3682531 /lib64/librt-2.13.so
3a22e07000-3a22e08000 rw-p 00007000 08:09 3682531 /lib64/librt-2.13.so
3a24c00000-3a24ce2000 r-xp 00000000 08:09 5248755 /usr/lib64/libgfortran.so.3.0.0
3a24ce2000-3a24ee2000 ---p 000e2000 08:09 5248755 /usr/lib64/libgfortran.so.3.0.0
3a24ee2000-3a24ee4000 rw-p 000e2000 08:09 5248755 /usr/lib64/libgfortran.so.3.0.0
3a24ee4000-3a24ee5000 rw-p 00000000 00:00 0
3a2f800000-3a2f80d000 r-xp 00000000 08:09 5284949 /usr/lib64/libgomp.so.1.0.0
3a2f80d000-3a2fa0c000 ---p 0000d000 08:09 5284949 /usr/lib64/libgomp.so.1.0.0
3a2fa0c000-3a2fa0d000 rw-p 0000c000 08:09 5284949 /usr/lib64/libgomp.so.1.0.0
7f98d0000000-7f98d0028000 rw-p 00000000 00:00 0
7f98d0028000-7f98d4000000 ---p 00000000 00:00 0
7f98d8000000-7f98d8028000 rw-p 00000000 00:00 0
7f98d8028000-7f98dc000000 ---p 00000000 00:00 0
7f98e0000000-7f98e0028000 rw-p 00000000 00:00 0
7f98e0028000-7f98e4000000 ---p 00000000 00:00 0
7f98e63e6000-7f98e6a44000 rw-p 00000000 00:00 0
7f98e6a44000-7f98e6a45000 ---p 00000000 00:00 0
7f98e6a45000-7f98e7245000 rw-p 00000000 00:00 0
7f98e7245000-7f98e7246000 ---p 00000000 00:00 0
7f98e7246000-7f98e7a46000 rw-p 00000000 00:00 0
7f98e7a46000-7f98e7a47000 ---p 00000000 00:00 0
7f98e7a47000-7f98e824c000 rw-p 00000000 00:00 0
7f98e8253000-7f98e8277000 rw-p 00000000 00:00 0
7fff5d402000-7fff5d43e000 rw-p 00000000 00:00 0 [stack]
7fff5d58b000-7fff5d58c000 r-xp 00000000 00:00 0 [vdso]
ffffffffff600000-ffffffffff601000 r-xp 00000000 00:00 0 [vsyscall]
Aborted (core dumped)

With debug options:
Model Input Parameters: ROMS/TOMS version 3.6
Monday - November 21, 2011 - 8:13:53 AM
-----------------------------------------------------------------------------
At line 1187 of file read_phypar.f90
Fortran runtime error: Index '0' of dimension 1 of array 'aout' below lower bound of 1

What cause this kind of problem?
Thanks in advance.

[kate]

At line 1187 of file read_phypar.f90
Fortran runtime error: Index '0' of dimension 1 of array 'aout' below lower bound of 1

What do you have on that line in your file?

[DTokarev]

Hi Kate,
This is the line from read_phypar.f90

CASE ('NC_DEFLATE')

Thanks

[kate]

Are you sure? That has neither Aout, nor any option to go out of bounds on any other array.

[m.hadfield]

Make sure you're looking at line 1187 of read_phypar.f90 (the preprocessed file) not read_phypar.F.

[DTokarev]

Dear Kate,
here lines 1181-1195 from Build/read_phypar.f90

CASE ('NUSER')
Npts=load_i(Nval, Rval, 1, Nuser)
CASE ('USER')
Npts=load_r(Nval, Rval, MAX(1,Nuser), user)
CASE ('NC_SHUFFLE')
Npts=load_i(Nval, Rval, 1, shuffle)
CASE ('NC_DEFLATE')
Npts=load_i(Nval, Rval, 1, deflate)
CASE ('NC_DLEVEL')
Npts=load_i(Nval, Rval, 1, deflate_level)
CASE ('GSTNAME')
label='GST - generalized stability theory analysis'
Npts=load_s1d(Nval, Cval, line, label, igrid, Nfiles, GST)

Dear m.hadfield,

I have checked 2 files: Build/read_phypar.f90 (see above) and ROMS/Utility/read_phypar.F (lines 2353-2365)

CASE ('NUSER')
Npts=load_i(Nval, Rval, 1, Nuser)
CASE ('USER')
Npts=load_r(Nval, Rval, MAX(1,Nuser), user)
CASE ('NC_SHUFFLE')
Npts=load_i(Nval, Rval, 1, shuffle)
CASE ('NC_DEFLATE')
Npts=load_i(Nval, Rval, 1, deflate)
CASE ('NC_DLEVEL')
Npts=load_i(Nval, Rval, 1, deflate_level)
CASE ('GSTNAME')
label='GST - generalized stability theory analysis'
Npts=load_s1d(Nval, Cval, line, label, igrid, Nfiles, GST)

and lines 1183-1188

CASE ('Hout(idVbcs)')
IF (idVbcs.eq.0) THEN
IF (Master) WRITE (out,280) 'idVbcs'
exit_flag=5
1187 RETURN
END IF

Thanks.

[arango]

I don't know what is the problem here. Please make sure that you are using an updated version of the Metadata file varinfo.dat. I have seen many users having this kind of problem in the past. They are always using a recent version of the code but and old version of varinfo.dat or ocean.in. Or introduced a typo when editing ocean.in. You should always use consistent versions of these files with your svn code version.

The error is not in Hout but in Aout. I don't know why the executable is giving you the incorrect source code line. This kind of error is typical when the index to Aout has not initialized yet. This is done in mod_ncparam.F when reading varinfo.dat.

The fact that this is just happening only to you tells me that it is very likely that you are using an inconsistent version of varinfo.dat.