[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[GEMPAK #EQC-771253]: dcgrib2 dumps core on GRIB2 GFS 1 deg grids



Robert,

Assuming that the nex2gini grash was at the point when the output file was being
compressed, that would follow that the libpng and or libz has a problem
with the compiler optimization. Was the 5.9.3 binary built recently
with the same compiler version as you are building 5.10.2, or
could the OS and system libraries have been updated since?

I see that there is an updated libpng 1-2.16 available which fixes
a pallette chunk bug in pre-1.2.12 (and we are using 1.2.8), so will plop
that in and then look at the compiler.

Steve Chiswell
Unidata User Support


> FYI..I copied over the 5.9.3 binary (built with -xO2) and it works fine..
> 
> 
> 
> -----Original Message-----
> From: Robert Mullenax
> Sent: Mon 4/16/2007 9:02 AM
> To: Robert Mullenax; address@hidden
> Cc: address@hidden; address@hidden
> Subject: RE: [GEMPAK #EQC-771253]: dcgrib2 dumps core on GRIB2 GFS 1 deg grids
> 
> I ran it by hand and it gets so far, then core dumps:
> 
> /usr2/config% /usr2/config/scripts/gempak/nex2gini.csh
> GRDAREA   Area covered by grid
> PROJ      Map projection/angles/margins|dr  MER
> KXKY      Number of grid points in x;y      10;10
> CPYFIL    Grid file whose navigation is to
> GFUNC     Scalar grid                       TMPC
> RADTIM    Radar composite current/dattim
> RADDUR    Radar time window (minutes prior
> RADFRQ    Update Frequency
> STNFIL    Station information file          SFSTNS.TBL
> RADMODE   Radar operational mode
> SATFIL    Satellite image filename(s)
> COMPRESS  Write output in compressed forma
> Parameters requested: GRDAREA,PROJ,KXKY,CPYFIL,GFUNC,RADTIM,RADDUR,RADFRQ,
> STNFIL,RADMODE,SATFIL,COMPRESS.
> GEMPAK-NEX2GINI> GEMPAK-NEX2GINI> GEMPAK-NEX2GINI> [NEX2GINI 0]  Using:   
> $RAD/NIDS/TLX/N0R/N0R_20070416_1359
> [NEX2GINI 0]  Using:   $RAD/NIDS/AMA/N0R/N0R_20070416_1358
> [NEX2GINI 0]  Using:   $RAD/NIDS/HGX/N0R/N0R_20070416_1358
> [NEX2GINI 0]  Using:   $RAD/NIDS/MLB/N0R/N0R_20070416_1356
> [NEX2GINI 0]  Using:   $RAD/NIDS/DDC/N0R/N0R_20070416_1353
> [NEX2GINI 0]  Using:   $RAD/NIDS/ICT/N0R/N0R_20070416_1357
> [NEX2GINI 0]  Using:   $RAD/NIDS/IWA/N0R/N0R_20070416_1355
> [NEX2GINI 0]  Using:   $RAD/NIDS/INX/N0R/N0R_20070416_1355
> [NEX2GINI 0]  Using:   $RAD/NIDS/LWX/N0R/N0R_20070416_1356
> [NEX2GINI 0]  Using:   $RAD/NIDS/LZK/N0R/N0R_20070416_1359
> [NEX2GINI 0]  Using:   $RAD/NIDS/FTG/N0R/N0R_20070416_1350
> [NEX2GINI 0]  Using:   $RAD/NIDS/LSX/N0R/N0R_20070416_1351
> [NEX2GINI 0]  Using:   $RAD/NIDS/LVX/N0R/N0R_20070416_1356
> [NEX2GINI 0]  Using:   $RAD/NIDS/DAX/N0R/N0R_20070416_1353
> [NEX2GINI 1]  Too old: $RAD/NIDS/BMX/N0R/N0R_20070416_1320
> [NEX2GINI 0]  Using:   $RAD/NIDS/BOX/N0R/N0R_20070416_1357
> [NEX2GINI 0]  Using:   $RAD/NIDS/FWS/N0R/N0R_20070416_1356
> [NEX2GINI 0]  Using:   $RAD/NIDS/UEX/N0R/N0R_20070416_1359
> [NEX2GINI 0]  Using:   $RAD/NIDS/VTX/N0R/N0R_20070416_1400
> [NEX2GINI 0]  Using:   $RAD/NIDS/LOT/N0R/N0R_20070416_1353
> [NEX2GINI 0]  Using:   $RAD/NIDS/TWX/N0R/N0R_20070416_1357
> [NEX2GINI 0]  Using:   $RAD/NIDS/CBX/N0R/N0R_20070416_1358
> Segmentation Fault (core dumped)
> No match
> 
> 
> 
> -----Original Message-----
> From: Robert Mullenax
> Sent: Mon 4/16/2007 8:57 AM
> To: address@hidden
> Cc: address@hidden; address@hidden
> Subject: RE: [GEMPAK #EQC-771253]: dcgrib2 dumps core on GRIB2 GFS 1 deg grids
> 
> 
> Steve,
> 
> As a follow up to this, nex2gini is also core dumping on my system (it waa 
> built with -xO2).  I have a script that runs every 10 minutes and in the past 
> 24 hours and hasn't ran because of a core dump about 70% of the time...oddly 
> enough it worked fine the other 30% of the time.
> 
> I hate to have to build that without optimization.  I may go back and copy 
> over the 5.9.3 binary.
> 
> 
> Thanks,
> Robert
> 
> -----Original Message-----
> From: Unidata GEMPAK Support [mailto:address@hidden]
> Sent: Fri 4/13/2007 6:16 PM
> To: Robert Mullenax
> Cc: address@hidden; address@hidden
> Subject: [GEMPAK #EQC-771253]: dcgrib2 dumps core on GRIB2 GFS 1 deg grids
> 
> Robert,
> 
> It may be that the jasper/jpeg 2000 libraries aren't playing well with the 
> compiler flag.
> That would limit the problem outside of the other programs (except nagrib2).
> 
> I will have to test the compiler options as affecting those 3rd party 
> libraries.
> 
> 
> Steve Chiswell
> Unidata User Support
> 
> 
> > It looks like the old Sun compiler optimization issue.  I was able to get 
> > it to work by first compiling GEMPAK without optimization and then setting 
> > the dcgrib2 binary aside, recompiling GEMPAK with -xO2, then copying the 
> > unoptimized dcgrib2 back into $GEMEXE,
> >
> > Hopefully there aren't any other GEMPAK programs affected.  I haven't found 
> > any.
> >
> > Thanks,
> > Robert Mullenax
> >
> >
> >
> > -----Original Message-----
> > From: Robert Mullenax
> > Sent: Thu 4/12/2007 8:25 PM
> > To: address@hidden
> > Subject: RE: [GEMPAK #EQC-771253]: dcgrib2 dumps core on GRIB2 GFS 1 deg 
> > grids
> >
> > Steve,
> >
> > Sorry I ran it with no -v just dcgrib2 filename and it coredumped.  I reran 
> > it using Unidata binary and -v 4.  Same thing.
> >
> > /usr2/config/temp% cat gfs.t12z.pgrbf00.grib2 | dcgrib2 -v 4 
> > YYYYMMDDHH_test.gem
> > Opening WMO Originating Center Table wmocenter.tbl...
> > Opening WMO GRIB2 Parameter Table g2varswmo2.tbl...
> > Opening WMO GRIB2 Vertical Coordinate Table g2vcrdwmo2.tbl...
> >
> > SECTION 0 - INDICATOR SECTION
> > IS byte        7 (discipline) = 0
> > IS byte        8 (edition)    = 2
> >
> > SECTION 1 - IDENTIFICATION SECTION
> > IDS bytes  6 -  7 (orig center)      = 7
> > IDS bytes  8 -  9 (sub center)       = 0
> > IDS byte       10 (Master table ver) = 2
> > IDS byte       11 (Local table ver)  = 1
> > IDS byte       12 (ref time sig.)    = 1
> > IDS bytes 13 - 14 (Year)             = 2007
> > IDS byte       15 (Month)            = 4
> > IDS byte       16 (Day)              = 12
> > IDS byte       17 (Hour)             = 12
> > IDS byte       18 (Minute)           = 0
> > IDS byte       19 (Second)           = 0
> > IDS byte       20 (Prod status)      = 0
> > IDS byte       21 (type of data)     = 1
> >
> > SECTION 3 - GRID DEFINITION SECTION
> > GDS byte        6 (src of grid def) = 0
> > GDS bytes  7 - 10 (no. of grid pts) = 65160
> > GDS byte       11 (nb per opt list) = 0
> > GDS byte       12 (interp opt list) = 0
> > Grid Definition Template 3.0:  6 0 0 0 0 0 0 360 181 0 0 90000000 0 48 
> > -90000000 359000000 1000000 1000000 0
> > GDS bytes 31 - 34 (Nx)            = 360
> > GDS bytes 35 - 38 (Ny)            = 181
> > GDS bytes 47 - 50 (La1)           = 90.000000
> > GDS bytes 51 - 54 (Lo1)           = 0.000000
> > GDS byte       55 (res flag)      = 48
> > GDS bytes 56 - 59 (La2)           = -90.000000
> > GDS bytes 60 - 63 (Lo2)           = 359.000000
> > GDS bytes 64 - 67 (Di)            = 1.000000
> > GDS bytes 68 - 71 (Dj)            = 1.000000
> > GDS byte       72 (mode)          = 0
> >
> > SECTION 4 - PRODUCT DEFINITION SECTION
> > PDS bytes  6 -  7 (no. of coords) = 0
> > Product Definition Template 4.0:  3 5 2 0 81 0 0 1 0 100 0 100000 255 0 0
> >
> > SECTION 5 - DATA REPRESENTATION SECTION
> > DRS bytes  6 -  9 (no. of data points) = 65160
> > Data Representation Template 5.40:  -980191232 0 1 14 0 0 255
> >
> > SECTION 6 - BIT-MAP SECTION
> > BMS byte        6 (bit-map indicator) = 255
> >
> > Segmentation fault (core dumped)
> > /usr2/config/temp%
> >
> >
> >
> >
> >
> > -----Original Message-----
> > From: Unidata GEMPAK Support [mailto:address@hidden]
> > Sent: Thu 4/12/2007 5:35 PM
> > To: Robert Mullenax
> > Cc: address@hidden
> > Subject: [GEMPAK #EQC-771253]: dcgrib2 dumps core on GRIB2 GFS 1 deg grids
> >
> > Robert,
> >
> > Your syntax for running the dc decoders is wrong. You need to use "-v #"
> > where # is a number from 1 to 4. Instead, you have the output file
> > template where the verbose level should go.
> >
> > Steve Chiswell
> > Unidata User Support
> >
> > >
> > > Steve,
> > >
> > > I downloaded and installed the Unidata 5.10.2 Solaris x86 binary and it 
> > > core dumps also.  Do you have a Solaris x86 box you can test it on?
> > >
> > > /usr/gempak/upc/GEMPAK5.10.2/os/x86/bin/dcgrib2
> > > /usr2/config/temp% cat gfs.t12z.pgrbf00.grib2 | dcgrib2 -v 
> > > YYYYMMDDHH_test.gem
> > > Opening WMO Originating Center Table wmocenter.tbl...
> > > Opening WMO GRIB2 Parameter Table g2varswmo2.tbl...
> > > Opening WMO GRIB2 Vertical Coordinate Table g2vcrdwmo2.tbl...
> > > Segmentation fault (core dumped)
> > > /usr2/config/temp% which dcgrib2
> > > /usr/gempak/upc/GEMPAK5.10.2/os/x86/bin/dcgrib2
> > > /usr2/config/temp% ls
> > > core                     logs/                    gfs.t12z.pgrbf00.grib2
> > >
> > >
> > > Thanks,
> > > Robert Mullenax
> > > CSBF Meteorology
> > >
> > >
> > > -----Original Message-----
> > > From: Unidata GEMPAK Support [mailto:address@hidden]
> > > Sent: Thu 4/12/2007 3:25 PM
> > > To: Robert Mullenax
> > > Cc: address@hidden
> > > Subject: [GEMPAK #EQC-771253]: dcgrib2 dumps core on GRIB2 GFS 1 deg grids
> > >
> > > Robert,
> > >
> > > Works for me, as it should since we are decoding all the 0.5 degree grib 
> > > 2 files
> > > from CONDUIT as well.
> > >
> > > Make sure you ftp in binary mode!
> > >
> > > ftp> bin
> > > 200 Type set to I
> > > ftp> get  gfs.t12z.pgrbf03.grib2
> > > local: gfs.t12z.pgrbf03.grib2 remote: gfs.t12z.pgrbf03.grib2
> > >
> > >
> > > > cat gfs.t12z.pgrbf03.grib2 | dcgrib2 -v 1 -d - YYYYMMDDHH_gfs.gem
> > > [12049493] 070412/1419[DC 3]  Starting up. Version 5.10.2
> > > Opening WMO Originating Center Table wmocenter.tbl...
> > > Opening WMO GRIB2 Parameter Table g2varswmo2.tbl...
> > > Opening WMO GRIB2 Vertical Coordinate Table g2vcrdwmo2.tbl...
> > > [12049493] 070412/1419[DCGRIB 0] Opened 2007041212_gfs.gem model:96 
> > > grid:255
> > > [12049493] 070412/1419[DECODE_GRIB2 0] HGHT [070412/1200F003] 1000:-1 
> > > PRES 360 181
> > > [12049493] 070412/1419[DECODE_GRIB2 0] HGHT [070412/1200F003] 975:-1 PRES 
> > > 360 181
> > > [12049493] 070412/1419[DECODE_GRIB2 0] HGHT [070412/1200F003] 950:-1 PRES 
> > > 360 181
> > > [12049493] 070412/1419[DECODE_GRIB2 0] HGHT [070412/1200F003] 925:-1 PRES 
> > > 360 181
> > > [12049493] 070412/1419[DECODE_GRIB2 0] HGHT [070412/1200F003] 900:-1 PRES 
> > > 360 181
> > >
> > > etc.....
> > >
> > > Steve Chiswell
> > > Unidata User Support
> > >
> > >
> > > >
> > > > I am trying to ftp GRIB2 format GFS 1 deg grib files and decode them 
> > > > from ftpprd, but GEMPAK 5.9.3 and 5.10.2 both core dump when I cat the 
> > > > downloaded file into dcgrib2.  Would it be possible for you to try to 
> > > > duplicate this?  The files can be retrieved from:
> > > >
> > > > ftpprd.ncep.noaa.gov
> > > >
> > > > /pub/data/nccf/com/gfs/prod/gfs........
> > > >
> > > > the file is e.g. gfs.t12z.pgrbf03.grib2
> > > >
> > > > Thanks,
> > > > Robert Mullenax
> > > > CSBF Meteorology
> > > >
> > > >
> > >
> > >
> > > Ticket Details
> > > ===================
> > > Ticket ID: EQC-771253
> > > Department: Support GEMPAK
> > > Priority: Normal
> > > Status: Closed
> > >
> > >
> > >
> > >
> >
> >
> > Ticket Details
> > ===================
> > Ticket ID: EQC-771253
> > Department: Support GEMPAK
> > Priority: Normal
> > Status: Closed
> >
> >
> >
> >
> >
> 
> 
> Ticket Details
> ===================
> Ticket ID: EQC-771253
> Department: Support GEMPAK
> Priority: Normal
> Status: Closed
> 
> 
> 
> 
> 
> 


Ticket Details
===================
Ticket ID: EQC-771253
Department: Support GEMPAK
Priority: Normal
Status: Closed