[Met_help] [rt.rap.ucar.edu #61830] History for NETCDF Issue
John Halley Gotway via RT
met_help at ucar.edu
Fri Jun 14 13:48:44 MDT 2013
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Hi,
Okay, now I've *really* got a problem with the compile that I can't solve.
Which version of NetCDF did you use to get METv4.1 to compile? The compile I am running here at AFWA failed for version 4.1.3 (4.1.3 works great for the compile of METv4.0 though), and also fails for the latest version of netcdf, 4.3.0. With the latter case, here's where the compilation fails:
*** Finished Making the vx_grid Library ***
make[3]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
make[3]: Entering directory `/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
/usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c -I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/compiled/include
In file included from nc_var_info.cc:24:
nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
The nectdf 'include' I compiled for 4.3.0 has only one file in it, netcdf.h, while 4.1.3 has a few more, including the netcdf.hh that adds the inclusion of one more file from the 4.1.3 include (netcdfcpp.h).
The routine say netcdf compiled successfully for me, but I had to disable a couple of things to get it to run correctly: (I ran the following: "./configure --prefix=/home/qcteam/netcdf/compiled --disable-netcdf-4 --disable-dap"). Is it possible one of those disables is the reason for the missing file(s)?
It looks as if the timing of the 4.1 release might be using an older netcdf than 4.3.0? Rather than try every iteration of netcdf between 4.1.3 and 4.3.0, I thought I'd ask. I wasn't going to update netcdf but the compile of 'pb2nc' while using 4.1.3 failed, suggesting I had to update.
Thanks,
Matt
// SIGNED //
Matthew C. Sittel
Associate Scientist
University Corporation for Atmospheric Research
16WS/WXN, Offutt AFB, NE
(402) 294-3473 DSN: 271-3473
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
From: John Halley Gotway
Time: Wed Jun 12 10:42:50 2013
Matt,
We've typically been compiling MET using NetCDF version 3 - the
specific version is netcdf-3.6.3.
However, I just downloaded netcdf-4.1.3 and compiled it with the GNU
compilers (gcc and g++) using the same options you did (--disable-
netcdf-4 --disable-dap). Then I recompiled METv4.1 using
netcdf-4.1.3, and it worked fine.
Your netcdf build definitely needs to have include/netcdf.hh - so that
why 4.3.0 wasn't working.
Can you please do a 'make clean' followed by 'make >& make_met.log'
when compiling with netcdf-4.1.3 and then send me the log file? What
type of compilers are you using? Also, it's possible that if
the version of the compiler has changed since the time you compiled
netcdf-4.1.3 to now, that could be causing a problem.
Thanks,
John
On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT wrote:
>
> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
> Queue: met_help
> Subject: NETCDF Issue
> Owner: Nobody
> Requestors: matthew.sittel.ctr at offutt.af.mil
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>
>
> Hi,
>
> Okay, now I've *really* got a problem with the compile that I can't
solve.
>
> Which version of NetCDF did you use to get METv4.1 to compile? The
compile I am running here at AFWA failed for version 4.1.3 (4.1.3
works great for the compile of METv4.0 though), and also fails for the
latest version of netcdf, 4.3.0. With the latter case, here's where
the compilation fails:
>
> *** Finished Making the vx_grid Library ***
>
> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/compiled/include
> In file included from nc_var_info.cc:24:
> nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
>
> The nectdf 'include' I compiled for 4.3.0 has only one file in it,
netcdf.h, while 4.1.3 has a few more, including the netcdf.hh that
adds the inclusion of one more file from the 4.1.3 include
(netcdfcpp.h).
>
> The routine say netcdf compiled successfully for me, but I had to
disable a couple of things to get it to run correctly: (I ran the
following: "./configure --prefix=/home/qcteam/netcdf/compiled
--disable-netcdf-4 --disable-dap"). Is it possible one of those
disables is the reason for the missing file(s)?
>
> It looks as if the timing of the 4.1 release might be using an older
netcdf than 4.3.0? Rather than try every iteration of netcdf between
4.1.3 and 4.3.0, I thought I'd ask. I wasn't going to update netcdf
but the compile of 'pb2nc' while using 4.1.3 failed, suggesting I had
to update.
>
> Thanks,
> Matt
>
> // SIGNED //
> Matthew C. Sittel
> Associate Scientist
> University Corporation for Atmospheric Research
> 16WS/WXN, Offutt AFB, NE
> (402) 294-3473 DSN: 271-3473
>
------------------------------------------------
Subject: NETCDF Issue
From: matthew.sittel.ctr at offutt.af.mil
Time: Wed Jun 12 11:00:33 2013
OK, let me try again using the 4.1.3 we know works (at least for
v4.0). Here's where it fails during the MET v4.1 build:
make[2]: Entering directory `/ibrix/home/qcteam/METv4.1/src/tools'
make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
/usr/bin/g++ -m32 ensemble_stat_conf_info.cc -Wall -Wshadow -static -g
-c -I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
-I/h/common/netcdf-4.1.3/include
/usr/bin/g++ -m32 -o ensemble_stat ensemble_stat.cc
ensemble_stat_conf_info.o \
-Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
-I/home/qcteam/METv4.1/include -I/h/common/netcdf-
4.1.3/include -I/h/common/gsl-1.15/include \
-L/home/qcteam/METv4.1/lib -L/h/common/netcdf-4.1.3/lib
-L/h/common/gsl-1.15/lib \
-lvx_stat_out \
-lvx_statistics \
-lvx_shapedata \
-lvx_gsl_prob \
-lvx_analysis_util \
-lvx_data2d_factory \
-lvx_data2d_nc_met \
-lvx_data2d_grib \
-lvx_data2d_nc_pinterp \
-lvx_data2d_nccf \
-lvx_data2d \
-lvx_nc_util \
-lvx_grid \
-lvx_config \
-lvx_cal \
-lvx_util \
-lvx_math \
-lvx_color \
-lvx_log \
-lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas \
/usr/bin/ld: skipping incompatible /h/common/netcdf-
4.1.3/lib/libnetcdf_c++.a when searching for -lnetcdf_c++
/usr/bin/ld: cannot find -lnetcdf_c++
collect2: ld returned 1 exit status
make[4]: *** [ensemble_stat] Error 1
make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
make[3]: *** [all] Error 2
make[3]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools/core'
make[2]: *** [all] Error 2
make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
make: *** [all] Error 2
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Wednesday, June 12, 2013 11:43 AM
To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
Matt,
We've typically been compiling MET using NetCDF version 3 - the
specific version is netcdf-3.6.3.
However, I just downloaded netcdf-4.1.3 and compiled it with the GNU
compilers (gcc and g++) using the same options you did (--disable-
netcdf-4 --disable-dap). Then I recompiled METv4.1 using netcdf-
4.1.3, and it worked fine.
Your netcdf build definitely needs to have include/netcdf.hh - so that
why 4.3.0 wasn't working.
Can you please do a 'make clean' followed by 'make >& make_met.log'
when compiling with netcdf-4.1.3 and then send me the log file? What
type of compilers are you using? Also, it's possible that if the
version of the compiler has changed since the time you compiled
netcdf-4.1.3 to now, that could be causing a problem.
Thanks,
John
On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT wrote:
>
> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
> Queue: met_help
> Subject: NETCDF Issue
> Owner: Nobody
> Requestors: matthew.sittel.ctr at offutt.af.mil
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830
> >
>
>
> Hi,
>
> Okay, now I've *really* got a problem with the compile that I can't
solve.
>
> Which version of NetCDF did you use to get METv4.1 to compile? The
compile I am running here at AFWA failed for version 4.1.3 (4.1.3
works great for the compile of METv4.0 though), and also fails for the
latest version of netcdf, 4.3.0. With the latter case, here's where
the compilation fails:
>
> *** Finished Making the vx_grid Library ***
>
> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/compiled/include
> In file included from nc_var_info.cc:24:
> nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
>
> The nectdf 'include' I compiled for 4.3.0 has only one file in it,
netcdf.h, while 4.1.3 has a few more, including the netcdf.hh that
adds the inclusion of one more file from the 4.1.3 include
(netcdfcpp.h).
>
> The routine say netcdf compiled successfully for me, but I had to
disable a couple of things to get it to run correctly: (I ran the
following: "./configure --prefix=/home/qcteam/netcdf/compiled
--disable-netcdf-4 --disable-dap"). Is it possible one of those
disables is the reason for the missing file(s)?
>
> It looks as if the timing of the 4.1 release might be using an older
netcdf than 4.3.0? Rather than try every iteration of netcdf between
4.1.3 and 4.3.0, I thought I'd ask. I wasn't going to update netcdf
but the compile of 'pb2nc' while using 4.1.3 failed, suggesting I had
to update.
>
> Thanks,
> Matt
>
> // SIGNED //
> Matthew C. Sittel
> Associate Scientist
> University Corporation for Atmospheric Research 16WS/WXN, Offutt
AFB,
> NE
> (402) 294-3473 DSN: 271-3473
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
From: John Halley Gotway
Time: Wed Jun 12 11:07:34 2013
Matt,
The answer is in the error message you sent me:
/usr/bin/ld: skipping incompatible /h/common/netcdf-
4.1.3/lib/libnetcdf_c++.a when searching for -lnetcdf_c++
The linker isn't using the version of the NetCDF library you told it
to because it is "incompatible". In my experience, that either means
that that library was built using a different family of
compilers (e.g. PGI vs GNU) or that it was built using an earlier
version of the GNU compilers. I suspect that if you tried to rebuild
METv4.0 right now with the current version of the GNU compilers,
you'd see the same linker error.
I'd suggest rebuilding the NetCDF library in /h/common/netcdf-4.1.3
using the current version of the GNU compilers and then try again.
Hope that helps.
Thanks,
John
On 06/12/2013 11:00 AM, matthew.sittel.ctr at offutt.af.mil via RT wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>
> OK, let me try again using the 4.1.3 we know works (at least for
v4.0). Here's where it fails during the MET v4.1 build:
>
> make[2]: Entering directory `/ibrix/home/qcteam/METv4.1/src/tools'
> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>
> /usr/bin/g++ -m32 ensemble_stat_conf_info.cc -Wall -Wshadow -static
-g -c -I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
-I/h/common/netcdf-4.1.3/include
> /usr/bin/g++ -m32 -o ensemble_stat ensemble_stat.cc
ensemble_stat_conf_info.o \
> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
> -I/home/qcteam/METv4.1/include -I/h/common/netcdf-
4.1.3/include -I/h/common/gsl-1.15/include \
> -L/home/qcteam/METv4.1/lib -L/h/common/netcdf-4.1.3/lib
-L/h/common/gsl-1.15/lib \
> -lvx_stat_out \
> -lvx_statistics \
> -lvx_shapedata \
> -lvx_gsl_prob \
> -lvx_analysis_util \
> -lvx_data2d_factory \
> -lvx_data2d_nc_met \
> -lvx_data2d_grib \
> -lvx_data2d_nc_pinterp \
> -lvx_data2d_nccf \
> -lvx_data2d \
> -lvx_nc_util \
> -lvx_grid \
> -lvx_config \
> -lvx_cal \
> -lvx_util \
> -lvx_math \
> -lvx_color \
> -lvx_log \
> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas \
>
> /usr/bin/ld: skipping incompatible /h/common/netcdf-
4.1.3/lib/libnetcdf_c++.a when searching for -lnetcdf_c++
> /usr/bin/ld: cannot find -lnetcdf_c++
> collect2: ld returned 1 exit status
> make[4]: *** [ensemble_stat] Error 1
> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
> make[3]: *** [all] Error 2
> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
> make[2]: *** [all] Error 2
> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
> make[1]: *** [all] Error 2
> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
> make: *** [all] Error 2
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Wednesday, June 12, 2013 11:43 AM
> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>
> Matt,
>
> We've typically been compiling MET using NetCDF version 3 - the
specific version is netcdf-3.6.3.
>
> However, I just downloaded netcdf-4.1.3 and compiled it with the GNU
compilers (gcc and g++) using the same options you did (--disable-
netcdf-4 --disable-dap). Then I recompiled METv4.1 using netcdf-
4.1.3, and it worked fine.
>
> Your netcdf build definitely needs to have include/netcdf.hh - so
that why 4.3.0 wasn't working.
>
> Can you please do a 'make clean' followed by 'make >& make_met.log'
when compiling with netcdf-4.1.3 and then send me the log file? What
type of compilers are you using? Also, it's possible that if the
version of the compiler has changed since the time you compiled
netcdf-4.1.3 to now, that could be causing a problem.
>
> Thanks,
> John
>
> On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>
>> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
>> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
>> Queue: met_help
>> Subject: NETCDF Issue
>> Owner: Nobody
>> Requestors: matthew.sittel.ctr at offutt.af.mil
>> Status: new
>> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830
>>>
>>
>>
>> Hi,
>>
>> Okay, now I've *really* got a problem with the compile that I can't
solve.
>>
>> Which version of NetCDF did you use to get METv4.1 to compile? The
compile I am running here at AFWA failed for version 4.1.3 (4.1.3
works great for the compile of METv4.0 though), and also fails for the
latest version of netcdf, 4.3.0. With the latter case, here's where
the compilation fails:
>>
>> *** Finished Making the vx_grid Library ***
>>
>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
>> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/compiled/include
>> In file included from nc_var_info.cc:24:
>> nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
>>
>> The nectdf 'include' I compiled for 4.3.0 has only one file in it,
netcdf.h, while 4.1.3 has a few more, including the netcdf.hh that
adds the inclusion of one more file from the 4.1.3 include
(netcdfcpp.h).
>>
>> The routine say netcdf compiled successfully for me, but I had to
disable a couple of things to get it to run correctly: (I ran the
following: "./configure --prefix=/home/qcteam/netcdf/compiled
--disable-netcdf-4 --disable-dap"). Is it possible one of those
disables is the reason for the missing file(s)?
>>
>> It looks as if the timing of the 4.1 release might be using an
older netcdf than 4.3.0? Rather than try every iteration of netcdf
between 4.1.3 and 4.3.0, I thought I'd ask. I wasn't going to update
netcdf but the compile of 'pb2nc' while using 4.1.3 failed, suggesting
I had to update.
>>
>> Thanks,
>> Matt
>>
>> // SIGNED //
>> Matthew C. Sittel
>> Associate Scientist
>> University Corporation for Atmospheric Research 16WS/WXN, Offutt
AFB,
>> NE
>> (402) 294-3473 DSN: 271-3473
>>
>
>
------------------------------------------------
Subject: NETCDF Issue
From: matthew.sittel.ctr at offutt.af.mil
Time: Wed Jun 12 12:11:45 2013
Hey John,
Okay, I re-compiled netcdf 4.1.3 successfully, and sourced it. Here’s
the error I get now. Since I get the same errors when compiling with
the older netcdf we already had installed and compiled in 2011, I
think we can rule out netcdf as the issue.
=====
make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
/usr/bin/g++ pb2nc_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include
/usr/bin/gfortran numpbmsg.f -Wall -Wshadow -static -ff2c -g -c
/usr/bin/gfortran openpb.f -Wall -Wshadow -static -ff2c -g -c
/usr/bin/gfortran closepb.f -Wall -Wshadow -static -ff2c -g -c
/usr/bin/gfortran readpb.f -Wall -Wshadow -static -ff2c -g -c
/usr/bin/gfortran dumppb.f -Wall -Wshadow -static -ff2c -g -c
/usr/bin/g++ -o pb2nc pb2nc.cc pb2nc_conf_info.o numpbmsg.o openpb.o
closepb.o readpb.o dumppb.o \
-Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
-I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/netcdf-
4.1.3/compiled/include -I/h/common/gsl-1.15/include
-I/h/common/BUFRLIB_v10-0-1 -I/h/common/libf2c \
-L/home/qcteam/METv4.1/lib -L/home/qcteam/netcdf/netcdf-
4.1.3/compiled/lib -L/h/common/gsl-1.15/lib -L/h/common/BUFRLIB_v10-0-
1 -L/h/common/libf2c \
-lvx_stat_out \
-lvx_statistics \
-lvx_shapedata \
-lvx_gsl_prob \
-lvx_analysis_util \
-lvx_data2d_grib \
-lvx_data2d \
-lvx_nc_util \
-lvx_grid \
-lvx_config \
-lvx_pb_util \
-lvx_cal \
-lvx_util \
-lvx_math \
-lvx_color \
-lvx_log \
-lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas -lbufr \
-lf2c -lgfortran
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:70: error:
\u2018::isalnum\u2019 has not been declared
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:71: error:
\u2018::isalpha\u2019 has not been declared
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:72: error:
\u2018::iscntrl\u2019 has not been declared
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:73: error:
\u2018::isdigit\u2019 has not been declared
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:74: error:
\u2018::isgraph\u2019 has not been declared
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:75: error:
\u2018::islower\u2019 has not been declared
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:76: error:
\u2018::isprint\u2019 has not been declared
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:77: error:
\u2018::ispunct\u2019 has not been declared
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:78: error:
\u2018::isspace\u2019 has not been declared
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:79: error:
\u2018::isupper\u2019 has not been declared
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:80: error:
\u2018::isxdigit\u2019 has not been declared
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:81: error:
\u2018::tolower\u2019 has not been declared
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:82: error:
\u2018::toupper\u2019 has not been declared
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:51: error: \u2018_ISupper\u2019 was not
declared in this scope
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:52: error: \u2018_ISlower\u2019 was not
declared in this scope
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:53: error: \u2018_ISalpha\u2019 was not
declared in this scope
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:54: error: \u2018_ISdigit\u2019 was not
declared in this scope
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:55: error: \u2018_ISxdigit\u2019 was not
declared in this scope
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:56: error: \u2018_ISspace\u2019 was not
declared in this scope
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:57: error: \u2018_ISprint\u2019 was not
declared in this scope
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:58: error: \u2018_ISalpha\u2019 was not
declared in this scope
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:58: error: \u2018_ISdigit\u2019 was not
declared in this scope
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:58: error: \u2018_ISpunct\u2019 was not
declared in this scope
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:59: error: \u2018_IScntrl\u2019 was not
declared in this scope
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:60: error: \u2018_ISpunct\u2019 was not
declared in this scope
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:61: error: \u2018_ISalpha\u2019 was not
declared in this scope
/usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:61: error: \u2018_ISdigit\u2019 was not
declared in this scope
make[4]: *** [pb2nc] Error 1
make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
make[3]: *** [all] Error 2
make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other'
make[2]: *** [all] Error 2
make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
make: *** [all] Error 2
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Wednesday, June 12, 2013 12:08 PM
To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
Matt,
The answer is in the error message you sent me:
/usr/bin/ld: skipping incompatible /h/common/netcdf-
4.1.3/lib/libnetcdf_c++.a when searching for -lnetcdf_c++
The linker isn't using the version of the NetCDF library you told it
to because it is "incompatible". In my experience, that either means
that that library was built using a different family of compilers
(e.g. PGI vs GNU) or that it was built using an earlier version of the
GNU compilers. I suspect that if you tried to rebuild METv4.0 right
now with the current version of the GNU compilers, you'd see the same
linker error.
I'd suggest rebuilding the NetCDF library in /h/common/netcdf-4.1.3
using the current version of the GNU compilers and then try again.
Hope that helps.
Thanks,
John
On 06/12/2013 11:00 AM, matthew.sittel.ctr at offutt.af.mil via RT wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>
> OK, let me try again using the 4.1.3 we know works (at least for
v4.0). Here's where it fails during the MET v4.1 build:
>
> make[2]: Entering directory `/ibrix/home/qcteam/METv4.1/src/tools'
> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>
> /usr/bin/g++ -m32 ensemble_stat_conf_info.cc -Wall -Wshadow -static
-g -c -I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
-I/h/common/netcdf-4.1.3/include
> /usr/bin/g++ -m32 -o ensemble_stat ensemble_stat.cc
ensemble_stat_conf_info.o \
> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
> -I/home/qcteam/METv4.1/include -I/h/common/netcdf-
4.1.3/include -I/h/common/gsl-1.15/include \
> -L/home/qcteam/METv4.1/lib -L/h/common/netcdf-4.1.3/lib
-L/h/common/gsl-1.15/lib \
> -lvx_stat_out \
> -lvx_statistics \
> -lvx_shapedata \
> -lvx_gsl_prob \
> -lvx_analysis_util \
> -lvx_data2d_factory \
> -lvx_data2d_nc_met \
> -lvx_data2d_grib \
> -lvx_data2d_nc_pinterp \
> -lvx_data2d_nccf \
> -lvx_data2d \
> -lvx_nc_util \
> -lvx_grid \
> -lvx_config \
> -lvx_cal \
> -lvx_util \
> -lvx_math \
> -lvx_color \
> -lvx_log \
> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas \
>
> /usr/bin/ld: skipping incompatible
> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
> -lnetcdf_c++
> /usr/bin/ld: cannot find -lnetcdf_c++
> collect2: ld returned 1 exit status
> make[4]: *** [ensemble_stat] Error 1
> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
> make[3]: *** [all] Error 2
> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
> make[2]: *** [all] Error 2
> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
> make[1]: *** [all] Error 2
> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
> make: *** [all] Error 2
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Wednesday, June 12, 2013 11:43 AM
> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>
> Matt,
>
> We've typically been compiling MET using NetCDF version 3 - the
specific version is netcdf-3.6.3.
>
> However, I just downloaded netcdf-4.1.3 and compiled it with the GNU
compilers (gcc and g++) using the same options you did (--disable-
netcdf-4 --disable-dap). Then I recompiled METv4.1 using netcdf-
4.1.3, and it worked fine.
>
> Your netcdf build definitely needs to have include/netcdf.hh - so
that why 4.3.0 wasn't working.
>
> Can you please do a 'make clean' followed by 'make >& make_met.log'
when compiling with netcdf-4.1.3 and then send me the log file? What
type of compilers are you using? Also, it's possible that if the
version of the compiler has changed since the time you compiled
netcdf-4.1.3 to now, that could be causing a problem.
>
> Thanks,
> John
>
> On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>
>> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
>> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
>> Queue: met_help
>> Subject: NETCDF Issue
>> Owner: Nobody
>> Requestors: matthew.sittel.ctr at offutt.af.mil
>> Status: new
>> Ticket <URL:
>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830
>>>
>>
>>
>> Hi,
>>
>> Okay, now I've *really* got a problem with the compile that I can't
solve.
>>
>> Which version of NetCDF did you use to get METv4.1 to compile? The
compile I am running here at AFWA failed for version 4.1.3 (4.1.3
works great for the compile of METv4.0 though), and also fails for the
latest version of netcdf, 4.3.0. With the latter case, here's where
the compilation fails:
>>
>> *** Finished Making the vx_grid Library ***
>>
>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
>> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/compiled/include
>> In file included from nc_var_info.cc:24:
>> nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
>>
>> The nectdf 'include' I compiled for 4.3.0 has only one file in it,
netcdf.h, while 4.1.3 has a few more, including the netcdf.hh that
adds the inclusion of one more file from the 4.1.3 include
(netcdfcpp.h).
>>
>> The routine say netcdf compiled successfully for me, but I had to
disable a couple of things to get it to run correctly: (I ran the
following: "./configure --prefix=/home/qcteam/netcdf/compiled
--disable-netcdf-4 --disable-dap"). Is it possible one of those
disables is the reason for the missing file(s)?
>>
>> It looks as if the timing of the 4.1 release might be using an
older netcdf than 4.3.0? Rather than try every iteration of netcdf
between 4.1.3 and 4.3.0, I thought I'd ask. I wasn't going to update
netcdf but the compile of 'pb2nc' while using 4.1.3 failed, suggesting
I had to update.
>>
>> Thanks,
>> Matt
>>
>> // SIGNED //
>> Matthew C. Sittel
>> Associate Scientist
>> University Corporation for Atmospheric Research 16WS/WXN, Offutt
AFB,
>> NE
>> (402) 294-3473 DSN: 271-3473
>>
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
From: John Halley Gotway
Time: Wed Jun 12 12:57:54 2013
Matt,
I read through some previous MET-Help questions and have a suggestion
for you to try - in user_defs.mk just empty out all of the "F2C_..."
settings:
F2C_BASE =
F2C_INCS =
F2C_LIBS =
F2C_LIBNAME =
Based on the log output you sent, it looks like you've specified
something in those settings. That's only necessary for the old GNU
g77, and you're using gfortran.
Please let me know if that does the trick.
Thanks,
John
On 06/12/2013 12:11 PM, matthew.sittel.ctr at offutt.af.mil via RT wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>
> Hey John,
>
> Okay, I re-compiled netcdf 4.1.3 successfully, and sourced it.
Here’s the error I get now. Since I get the same errors when
compiling with the older netcdf we already had installed and compiled
in 2011, I think we can rule out netcdf as the issue.
>
> =====
>
> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>
> /usr/bin/g++ pb2nc_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include
>
> /usr/bin/gfortran numpbmsg.f -Wall -Wshadow -static -ff2c -g -c
>
> /usr/bin/gfortran openpb.f -Wall -Wshadow -static -ff2c -g -c
>
> /usr/bin/gfortran closepb.f -Wall -Wshadow -static -ff2c -g -c
>
> /usr/bin/gfortran readpb.f -Wall -Wshadow -static -ff2c -g -c
>
> /usr/bin/gfortran dumppb.f -Wall -Wshadow -static -ff2c -g -c
> /usr/bin/g++ -o pb2nc pb2nc.cc pb2nc_conf_info.o numpbmsg.o
openpb.o closepb.o readpb.o dumppb.o \
> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
> -I/home/qcteam/METv4.1/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include -I/h/common/gsl-
1.15/include -I/h/common/BUFRLIB_v10-0-1 -I/h/common/libf2c \
> -L/home/qcteam/METv4.1/lib -L/home/qcteam/netcdf/netcdf-
4.1.3/compiled/lib -L/h/common/gsl-1.15/lib -L/h/common/BUFRLIB_v10-0-
1 -L/h/common/libf2c \
> -lvx_stat_out \
> -lvx_statistics \
> -lvx_shapedata \
> -lvx_gsl_prob \
> -lvx_analysis_util \
> -lvx_data2d_grib \
> -lvx_data2d \
> -lvx_nc_util \
> -lvx_grid \
> -lvx_config \
> -lvx_pb_util \
> -lvx_cal \
> -lvx_util \
> -lvx_math \
> -lvx_color \
> -lvx_log \
> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas -lbufr \
> -lf2c -lgfortran
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:70: error:
\u2018::isalnum\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:71: error:
\u2018::isalpha\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:72: error:
\u2018::iscntrl\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:73: error:
\u2018::isdigit\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:74: error:
\u2018::isgraph\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:75: error:
\u2018::islower\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:76: error:
\u2018::isprint\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:77: error:
\u2018::ispunct\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:78: error:
\u2018::isspace\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:79: error:
\u2018::isupper\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:80: error:
\u2018::isxdigit\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:81: error:
\u2018::tolower\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/cctype:82: error:
\u2018::toupper\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:51: error: \u2018_ISupper\u2019 was not
declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:52: error: \u2018_ISlower\u2019 was not
declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:53: error: \u2018_ISalpha\u2019 was not
declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:54: error: \u2018_ISdigit\u2019 was not
declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:55: error: \u2018_ISxdigit\u2019 was not
declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:56: error: \u2018_ISspace\u2019 was not
declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:57: error: \u2018_ISprint\u2019 was not
declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:58: error: \u2018_ISalpha\u2019 was not
declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:58: error: \u2018_ISdigit\u2019 was not
declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:58: error: \u2018_ISpunct\u2019 was not
declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:59: error: \u2018_IScntrl\u2019 was not
declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:60: error: \u2018_ISpunct\u2019 was not
declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:61: error: \u2018_ISalpha\u2019 was not
declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x86_64-redhat-
linux/bits/ctype_base.h:61: error: \u2018_ISdigit\u2019 was not
declared in this scope
> make[4]: *** [pb2nc] Error 1
> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
> make[3]: *** [all] Error 2
> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other'
> make[2]: *** [all] Error 2
> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
> make[1]: *** [all] Error 2
> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
> make: *** [all] Error 2
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Wednesday, June 12, 2013 12:08 PM
> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>
> Matt,
>
> The answer is in the error message you sent me:
> /usr/bin/ld: skipping incompatible /h/common/netcdf-
4.1.3/lib/libnetcdf_c++.a when searching for -lnetcdf_c++
>
> The linker isn't using the version of the NetCDF library you told it
to because it is "incompatible". In my experience, that either means
that that library was built using a different family of compilers
(e.g. PGI vs GNU) or that it was built using an earlier version of the
GNU compilers. I suspect that if you tried to rebuild METv4.0 right
now with the current version of the GNU compilers, you'd see the same
linker error.
>
> I'd suggest rebuilding the NetCDF library in /h/common/netcdf-4.1.3
using the current version of the GNU compilers and then try again.
>
> Hope that helps.
>
> Thanks,
> John
>
> On 06/12/2013 11:00 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>
>> OK, let me try again using the 4.1.3 we know works (at least for
v4.0). Here's where it fails during the MET v4.1 build:
>>
>> make[2]: Entering directory `/ibrix/home/qcteam/METv4.1/src/tools'
>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>
>> /usr/bin/g++ -m32 ensemble_stat_conf_info.cc -Wall -Wshadow -static
-g -c -I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
-I/h/common/netcdf-4.1.3/include
>> /usr/bin/g++ -m32 -o ensemble_stat ensemble_stat.cc
ensemble_stat_conf_info.o \
>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>> -I/home/qcteam/METv4.1/include -I/h/common/netcdf-
4.1.3/include -I/h/common/gsl-1.15/include \
>> -L/home/qcteam/METv4.1/lib -L/h/common/netcdf-4.1.3/lib
-L/h/common/gsl-1.15/lib \
>> -lvx_stat_out \
>> -lvx_statistics \
>> -lvx_shapedata \
>> -lvx_gsl_prob \
>> -lvx_analysis_util \
>> -lvx_data2d_factory \
>> -lvx_data2d_nc_met \
>> -lvx_data2d_grib \
>> -lvx_data2d_nc_pinterp \
>> -lvx_data2d_nccf \
>> -lvx_data2d \
>> -lvx_nc_util \
>> -lvx_grid \
>> -lvx_config \
>> -lvx_cal \
>> -lvx_util \
>> -lvx_math \
>> -lvx_color \
>> -lvx_log \
>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas \
>>
>> /usr/bin/ld: skipping incompatible
>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>> -lnetcdf_c++
>> /usr/bin/ld: cannot find -lnetcdf_c++
>> collect2: ld returned 1 exit status
>> make[4]: *** [ensemble_stat] Error 1
>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>> make[3]: *** [all] Error 2
>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>> make[2]: *** [all] Error 2
>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>> make[1]: *** [all] Error 2
>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>> make: *** [all] Error 2
>>
>> -----Original Message-----
>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>> Sent: Wednesday, June 12, 2013 11:43 AM
>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>
>> Matt,
>>
>> We've typically been compiling MET using NetCDF version 3 - the
specific version is netcdf-3.6.3.
>>
>> However, I just downloaded netcdf-4.1.3 and compiled it with the
GNU compilers (gcc and g++) using the same options you did (--disable-
netcdf-4 --disable-dap). Then I recompiled METv4.1 using netcdf-
4.1.3, and it worked fine.
>>
>> Your netcdf build definitely needs to have include/netcdf.hh - so
that why 4.3.0 wasn't working.
>>
>> Can you please do a 'make clean' followed by 'make >& make_met.log'
when compiling with netcdf-4.1.3 and then send me the log file? What
type of compilers are you using? Also, it's possible that if the
version of the compiler has changed since the time you compiled
netcdf-4.1.3 to now, that could be causing a problem.
>>
>> Thanks,
>> John
>>
>> On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>
>>> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
>>> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
>>> Queue: met_help
>>> Subject: NETCDF Issue
>>> Owner: Nobody
>>> Requestors: matthew.sittel.ctr at offutt.af.mil
>>> Status: new
>>> Ticket <URL:
>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830
>>>>
>>>
>>>
>>> Hi,
>>>
>>> Okay, now I've *really* got a problem with the compile that I
can't solve.
>>>
>>> Which version of NetCDF did you use to get METv4.1 to compile?
The compile I am running here at AFWA failed for version 4.1.3 (4.1.3
works great for the compile of METv4.0 though), and also fails for the
latest version of netcdf, 4.3.0. With the latter case, here's where
the compilation fails:
>>>
>>> *** Finished Making the vx_grid Library ***
>>>
>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
>>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
>>> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/compiled/include
>>> In file included from nc_var_info.cc:24:
>>> nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
>>>
>>> The nectdf 'include' I compiled for 4.3.0 has only one file in it,
netcdf.h, while 4.1.3 has a few more, including the netcdf.hh that
adds the inclusion of one more file from the 4.1.3 include
(netcdfcpp.h).
>>>
>>> The routine say netcdf compiled successfully for me, but I had to
disable a couple of things to get it to run correctly: (I ran the
following: "./configure --prefix=/home/qcteam/netcdf/compiled
--disable-netcdf-4 --disable-dap"). Is it possible one of those
disables is the reason for the missing file(s)?
>>>
>>> It looks as if the timing of the 4.1 release might be using an
older netcdf than 4.3.0? Rather than try every iteration of netcdf
between 4.1.3 and 4.3.0, I thought I'd ask. I wasn't going to update
netcdf but the compile of 'pb2nc' while using 4.1.3 failed, suggesting
I had to update.
>>>
>>> Thanks,
>>> Matt
>>>
>>> // SIGNED //
>>> Matthew C. Sittel
>>> Associate Scientist
>>> University Corporation for Atmospheric Research 16WS/WXN, Offutt
AFB,
>>> NE
>>> (402) 294-3473 DSN: 271-3473
>>>
>>
>>
>
>
------------------------------------------------
Subject: NETCDF Issue
From: matthew.sittel.ctr at offutt.af.mil
Time: Wed Jun 12 13:26:33 2013
I got through that step after removing the F2C settings, as you
suggested.
I got everything else to compile *if* I left out modis_regrid and the
tropical steps (i.e., set those last two options to 0 in the mk file).
Modis mentioned something about "hdf", which I think is part of the
NetCDF routine-I recall reading that in the "INSTALL" part of NetCDF.
The error I got from the tropical part might be useful to you?
===
make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
/usr/bin/gfortran acerr.f -Wall -Wshadow -static -ff2c -g -c
/usr/bin/gfortran clip_shif.f -Wall -Wshadow -static -ff2c -g -c >
/dev/null 2>&1
/usr/bin/g++ tc_pairs_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
/usr/bin/g++ -o tc_pairs tc_pairs.cc acerr.o clip_shif.o
tc_pairs_conf_info.o \
-Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
-I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/netcdf-
4.1.3/compiled/include -I/h/common/gsl-1.15/include \
-L/home/qcteam/METv4.1/lib -L/home/qcteam/netcdf/netcdf-
4.1.3/compiled/lib -L/h/common/gsl-1.15/lib \
-lvx_config \
-lvx_data2d_nc_met \
-lvx_analysis_util \
-lvx_tc_util \
-lvx_nc_util \
-lvx_grid \
-lvx_config \
-lvx_util \
-lvx_math \
-lvx_cal \
-lvx_log \
-lm \
-lnetcdf_c++ -lnetcdf \
-lgfortran
/tmp/ccJGmQPZ.o: In function `derive_baseline_model':
/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs/tc_pairs.cc:1229:
undefined reference to `oclipd5_test_'
collect2: ld returned 1 exit status
make[4]: *** [tc_pairs] Error 1
make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
make[3]: *** [all] Error 2
make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils'
make[2]: *** [all] Error 2
make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
make: *** [all] Error 2
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Wednesday, June 12, 2013 1:58 PM
To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
Matt,
I read through some previous MET-Help questions and have a suggestion
for you to try - in user_defs.mk just empty out all of the "F2C_..."
settings:
F2C_BASE =
F2C_INCS =
F2C_LIBS =
F2C_LIBNAME =
Based on the log output you sent, it looks like you've specified
something in those settings. That's only necessary for the old GNU
g77, and you're using gfortran.
Please let me know if that does the trick.
Thanks,
John
On 06/12/2013 12:11 PM, matthew.sittel.ctr at offutt.af.mil via RT wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>
> Hey John,
>
> Okay, I re-compiled netcdf 4.1.3 successfully, and sourced it.
Here’s the error I get now. Since I get the same errors when
compiling with the older netcdf we already had installed and compiled
in 2011, I think we can rule out netcdf as the issue.
>
> =====
>
> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>
> /usr/bin/g++ pb2nc_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include
>
> /usr/bin/gfortran numpbmsg.f -Wall -Wshadow -static -ff2c -g -c
>
> /usr/bin/gfortran openpb.f -Wall -Wshadow -static -ff2c -g -c
>
> /usr/bin/gfortran closepb.f -Wall -Wshadow -static -ff2c -g -c
>
> /usr/bin/gfortran readpb.f -Wall -Wshadow -static -ff2c -g -c
>
> /usr/bin/gfortran dumppb.f -Wall -Wshadow -static -ff2c -g -c
> /usr/bin/g++ -o pb2nc pb2nc.cc pb2nc_conf_info.o numpbmsg.o
openpb.o closepb.o readpb.o dumppb.o \
> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
> -I/home/qcteam/METv4.1/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include -I/h/common/gsl-
1.15/include -I/h/common/BUFRLIB_v10-0-1 -I/h/common/libf2c \
> -L/home/qcteam/METv4.1/lib -L/home/qcteam/netcdf/netcdf-
4.1.3/compiled/lib -L/h/common/gsl-1.15/lib -L/h/common/BUFRLIB_v10-0-
1 -L/h/common/libf2c \
> -lvx_stat_out \
> -lvx_statistics \
> -lvx_shapedata \
> -lvx_gsl_prob \
> -lvx_analysis_util \
> -lvx_data2d_grib \
> -lvx_data2d \
> -lvx_nc_util \
> -lvx_grid \
> -lvx_config \
> -lvx_pb_util \
> -lvx_cal \
> -lvx_util \
> -lvx_math \
> -lvx_color \
> -lvx_log \
> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas -lbufr \
> -lf2c -lgfortran
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
> ctype:70: error: \u2018::isalnum\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
> ctype:71: error: \u2018::isalpha\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
> ctype:72: error: \u2018::iscntrl\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
> ctype:73: error: \u2018::isdigit\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
> ctype:74: error: \u2018::isgraph\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
> ctype:75: error: \u2018::islower\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
> ctype:76: error: \u2018::isprint\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
> ctype:77: error: \u2018::ispunct\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
> ctype:78: error: \u2018::isspace\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
> ctype:79: error: \u2018::isupper\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
> ctype:80: error: \u2018::isxdigit\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
> ctype:81: error: \u2018::tolower\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
> ctype:82: error: \u2018::toupper\u2019 has not been declared
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:51: error: \u2018_ISupper\u2019
> was not declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:52: error: \u2018_ISlower\u2019
> was not declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:53: error: \u2018_ISalpha\u2019
> was not declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:54: error: \u2018_ISdigit\u2019
> was not declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:55: error:
\u2018_ISxdigit\u2019
> was not declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:56: error: \u2018_ISspace\u2019
> was not declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:57: error: \u2018_ISprint\u2019
> was not declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:58: error: \u2018_ISalpha\u2019
> was not declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:58: error: \u2018_ISdigit\u2019
> was not declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:58: error: \u2018_ISpunct\u2019
> was not declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:59: error: \u2018_IScntrl\u2019
> was not declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:60: error: \u2018_ISpunct\u2019
> was not declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:61: error: \u2018_ISalpha\u2019
> was not declared in this scope
> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
> 86_64-redhat-linux/bits/ctype_base.h:61: error: \u2018_ISdigit\u2019
> was not declared in this scope
> make[4]: *** [pb2nc] Error 1
> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
> make[3]: *** [all] Error 2
> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other'
> make[2]: *** [all] Error 2
> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
> make[1]: *** [all] Error 2
> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
> make: *** [all] Error 2
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Wednesday, June 12, 2013 12:08 PM
> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>
> Matt,
>
> The answer is in the error message you sent me:
> /usr/bin/ld: skipping incompatible
> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
> -lnetcdf_c++
>
> The linker isn't using the version of the NetCDF library you told it
to because it is "incompatible". In my experience, that either means
that that library was built using a different family of compilers
(e.g. PGI vs GNU) or that it was built using an earlier version of the
GNU compilers. I suspect that if you tried to rebuild METv4.0 right
now with the current version of the GNU compilers, you'd see the same
linker error.
>
> I'd suggest rebuilding the NetCDF library in /h/common/netcdf-4.1.3
using the current version of the GNU compilers and then try again.
>
> Hope that helps.
>
> Thanks,
> John
>
> On 06/12/2013 11:00 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>
>> OK, let me try again using the 4.1.3 we know works (at least for
v4.0). Here's where it fails during the MET v4.1 build:
>>
>> make[2]: Entering directory `/ibrix/home/qcteam/METv4.1/src/tools'
>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>
>> /usr/bin/g++ -m32 ensemble_stat_conf_info.cc -Wall -Wshadow -static
-g -c -I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
-I/h/common/netcdf-4.1.3/include
>> /usr/bin/g++ -m32 -o ensemble_stat ensemble_stat.cc
ensemble_stat_conf_info.o \
>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>> -I/home/qcteam/METv4.1/include -I/h/common/netcdf-
4.1.3/include -I/h/common/gsl-1.15/include \
>> -L/home/qcteam/METv4.1/lib -L/h/common/netcdf-4.1.3/lib
-L/h/common/gsl-1.15/lib \
>> -lvx_stat_out \
>> -lvx_statistics \
>> -lvx_shapedata \
>> -lvx_gsl_prob \
>> -lvx_analysis_util \
>> -lvx_data2d_factory \
>> -lvx_data2d_nc_met \
>> -lvx_data2d_grib \
>> -lvx_data2d_nc_pinterp \
>> -lvx_data2d_nccf \
>> -lvx_data2d \
>> -lvx_nc_util \
>> -lvx_grid \
>> -lvx_config \
>> -lvx_cal \
>> -lvx_util \
>> -lvx_math \
>> -lvx_color \
>> -lvx_log \
>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas \
>>
>> /usr/bin/ld: skipping incompatible
>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>> -lnetcdf_c++
>> /usr/bin/ld: cannot find -lnetcdf_c++
>> collect2: ld returned 1 exit status
>> make[4]: *** [ensemble_stat] Error 1
>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>> make[3]: *** [all] Error 2
>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>> make[2]: *** [all] Error 2
>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>> make[1]: *** [all] Error 2
>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>> make: *** [all] Error 2
>>
>> -----Original Message-----
>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>> Sent: Wednesday, June 12, 2013 11:43 AM
>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>
>> Matt,
>>
>> We've typically been compiling MET using NetCDF version 3 - the
specific version is netcdf-3.6.3.
>>
>> However, I just downloaded netcdf-4.1.3 and compiled it with the
GNU compilers (gcc and g++) using the same options you did (--disable-
netcdf-4 --disable-dap). Then I recompiled METv4.1 using netcdf-
4.1.3, and it worked fine.
>>
>> Your netcdf build definitely needs to have include/netcdf.hh - so
that why 4.3.0 wasn't working.
>>
>> Can you please do a 'make clean' followed by 'make >& make_met.log'
when compiling with netcdf-4.1.3 and then send me the log file? What
type of compilers are you using? Also, it's possible that if the
version of the compiler has changed since the time you compiled
netcdf-4.1.3 to now, that could be causing a problem.
>>
>> Thanks,
>> John
>>
>> On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>
>>> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
>>> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
>>> Queue: met_help
>>> Subject: NETCDF Issue
>>> Owner: Nobody
>>> Requestors: matthew.sittel.ctr at offutt.af.mil
>>> Status: new
>>> Ticket <URL:
>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830
>>>>
>>>
>>>
>>> Hi,
>>>
>>> Okay, now I've *really* got a problem with the compile that I
can't solve.
>>>
>>> Which version of NetCDF did you use to get METv4.1 to compile?
The compile I am running here at AFWA failed for version 4.1.3 (4.1.3
works great for the compile of METv4.0 though), and also fails for the
latest version of netcdf, 4.3.0. With the latter case, here's where
the compilation fails:
>>>
>>> *** Finished Making the vx_grid Library ***
>>>
>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
>>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
>>> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/compiled/include
>>> In file included from nc_var_info.cc:24:
>>> nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
>>>
>>> The nectdf 'include' I compiled for 4.3.0 has only one file in it,
netcdf.h, while 4.1.3 has a few more, including the netcdf.hh that
adds the inclusion of one more file from the 4.1.3 include
(netcdfcpp.h).
>>>
>>> The routine say netcdf compiled successfully for me, but I had to
disable a couple of things to get it to run correctly: (I ran the
following: "./configure --prefix=/home/qcteam/netcdf/compiled
--disable-netcdf-4 --disable-dap"). Is it possible one of those
disables is the reason for the missing file(s)?
>>>
>>> It looks as if the timing of the 4.1 release might be using an
older netcdf than 4.3.0? Rather than try every iteration of netcdf
between 4.1.3 and 4.3.0, I thought I'd ask. I wasn't going to update
netcdf but the compile of 'pb2nc' while using 4.1.3 failed, suggesting
I had to update.
>>>
>>> Thanks,
>>> Matt
>>>
>>> // SIGNED //
>>> Matthew C. Sittel
>>> Associate Scientist
>>> University Corporation for Atmospheric Research 16WS/WXN, Offutt
>>> AFB, NE
>>> (402) 294-3473 DSN: 271-3473
>>>
>>
>>
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
From: John Halley Gotway
Time: Wed Jun 12 14:01:22 2013
Matt,
If you don't have the HDF and HDF-EOS libraries set up, the MODIS
piece won't compile.
Regarding the problem with compiling tc_utils, I think you're missing
the "-fno-second-underscore" option in the FC_FLAGS line. It's
included in the default user_defs_gnu.mk, but for some reason it
appears not to be present in yours. Please try adding that back in,
enable the compilation of the tc_utils, and try again.
Thanks,
John
On 06/12/2013 01:26 PM, matthew.sittel.ctr at offutt.af.mil via RT wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>
> I got through that step after removing the F2C settings, as you
suggested.
>
> I got everything else to compile *if* I left out modis_regrid and
the tropical steps (i.e., set those last two options to 0 in the mk
file).
>
> Modis mentioned something about "hdf", which I think is part of the
NetCDF routine-I recall reading that in the "INSTALL" part of NetCDF.
>
> The error I got from the tropical part might be useful to you?
>
> ===
> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
>
> /usr/bin/gfortran acerr.f -Wall -Wshadow -static -ff2c -g -c
>
> /usr/bin/gfortran clip_shif.f -Wall -Wshadow -static -ff2c -g -c >
/dev/null 2>&1
>
> /usr/bin/g++ tc_pairs_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>
> /usr/bin/g++ -o tc_pairs tc_pairs.cc acerr.o clip_shif.o
tc_pairs_conf_info.o \
> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
> -I/home/qcteam/METv4.1/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include -I/h/common/gsl-
1.15/include \
> -L/home/qcteam/METv4.1/lib -L/home/qcteam/netcdf/netcdf-
4.1.3/compiled/lib -L/h/common/gsl-1.15/lib \
> -lvx_config \
> -lvx_data2d_nc_met \
> -lvx_analysis_util \
> -lvx_tc_util \
> -lvx_nc_util \
> -lvx_grid \
> -lvx_config \
> -lvx_util \
> -lvx_math \
> -lvx_cal \
> -lvx_log \
> -lm \
> -lnetcdf_c++ -lnetcdf \
> -lgfortran
> /tmp/ccJGmQPZ.o: In function `derive_baseline_model':
> /home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs/tc_pairs.cc:1229:
undefined reference to `oclipd5_test_'
> collect2: ld returned 1 exit status
> make[4]: *** [tc_pairs] Error 1
> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
> make[3]: *** [all] Error 2
> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils'
> make[2]: *** [all] Error 2
> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
> make[1]: *** [all] Error 2
> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
> make: *** [all] Error 2
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Wednesday, June 12, 2013 1:58 PM
> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>
> Matt,
>
> I read through some previous MET-Help questions and have a
suggestion for you to try - in user_defs.mk just empty out all of the
"F2C_..." settings:
> F2C_BASE =
> F2C_INCS =
> F2C_LIBS =
> F2C_LIBNAME =
>
> Based on the log output you sent, it looks like you've specified
something in those settings. That's only necessary for the old GNU
g77, and you're using gfortran.
>
> Please let me know if that does the trick.
>
> Thanks,
> John
>
> On 06/12/2013 12:11 PM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>
>> Hey John,
>>
>> Okay, I re-compiled netcdf 4.1.3 successfully, and sourced it.
Here’s the error I get now. Since I get the same errors when
compiling with the older netcdf we already had installed and compiled
in 2011, I think we can rule out netcdf as the issue.
>>
>> =====
>>
>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>>
>> /usr/bin/g++ pb2nc_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include
>>
>> /usr/bin/gfortran numpbmsg.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran openpb.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran closepb.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran readpb.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran dumppb.f -Wall -Wshadow -static -ff2c -g -c
>> /usr/bin/g++ -o pb2nc pb2nc.cc pb2nc_conf_info.o numpbmsg.o
openpb.o closepb.o readpb.o dumppb.o \
>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>> -I/home/qcteam/METv4.1/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include -I/h/common/gsl-
1.15/include -I/h/common/BUFRLIB_v10-0-1 -I/h/common/libf2c \
>> -L/home/qcteam/METv4.1/lib -L/home/qcteam/netcdf/netcdf-
4.1.3/compiled/lib -L/h/common/gsl-1.15/lib -L/h/common/BUFRLIB_v10-0-
1 -L/h/common/libf2c \
>> -lvx_stat_out \
>> -lvx_statistics \
>> -lvx_shapedata \
>> -lvx_gsl_prob \
>> -lvx_analysis_util \
>> -lvx_data2d_grib \
>> -lvx_data2d \
>> -lvx_nc_util \
>> -lvx_grid \
>> -lvx_config \
>> -lvx_pb_util \
>> -lvx_cal \
>> -lvx_util \
>> -lvx_math \
>> -lvx_color \
>> -lvx_log \
>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas -lbufr \
>> -lf2c -lgfortran
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
>> ctype:70: error: \u2018::isalnum\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
>> ctype:71: error: \u2018::isalpha\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
>> ctype:72: error: \u2018::iscntrl\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
>> ctype:73: error: \u2018::isdigit\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
>> ctype:74: error: \u2018::isgraph\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
>> ctype:75: error: \u2018::islower\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
>> ctype:76: error: \u2018::isprint\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
>> ctype:77: error: \u2018::ispunct\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
>> ctype:78: error: \u2018::isspace\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
>> ctype:79: error: \u2018::isupper\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
>> ctype:80: error: \u2018::isxdigit\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
>> ctype:81: error: \u2018::tolower\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/c
>> ctype:82: error: \u2018::toupper\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:51: error:
\u2018_ISupper\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:52: error:
\u2018_ISlower\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:53: error:
\u2018_ISalpha\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:54: error:
\u2018_ISdigit\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:55: error:
\u2018_ISxdigit\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:56: error:
\u2018_ISspace\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:57: error:
\u2018_ISprint\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISalpha\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISdigit\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISpunct\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:59: error:
\u2018_IScntrl\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:60: error:
\u2018_ISpunct\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
\u2018_ISalpha\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/x
>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
\u2018_ISdigit\u2019
>> was not declared in this scope
>> make[4]: *** [pb2nc] Error 1
>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>> make[3]: *** [all] Error 2
>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other'
>> make[2]: *** [all] Error 2
>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>> make[1]: *** [all] Error 2
>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>> make: *** [all] Error 2
>>
>> -----Original Message-----
>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>> Sent: Wednesday, June 12, 2013 12:08 PM
>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>
>> Matt,
>>
>> The answer is in the error message you sent me:
>> /usr/bin/ld: skipping incompatible
>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>> -lnetcdf_c++
>>
>> The linker isn't using the version of the NetCDF library you told
it to because it is "incompatible". In my experience, that either
means that that library was built using a different family of
compilers (e.g. PGI vs GNU) or that it was built using an earlier
version of the GNU compilers. I suspect that if you tried to rebuild
METv4.0 right now with the current version of the GNU compilers, you'd
see the same linker error.
>>
>> I'd suggest rebuilding the NetCDF library in /h/common/netcdf-4.1.3
using the current version of the GNU compilers and then try again.
>>
>> Hope that helps.
>>
>> Thanks,
>> John
>>
>> On 06/12/2013 11:00 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>
>>> OK, let me try again using the 4.1.3 we know works (at least for
v4.0). Here's where it fails during the MET v4.1 build:
>>>
>>> make[2]: Entering directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>>
>>> /usr/bin/g++ -m32 ensemble_stat_conf_info.cc -Wall -Wshadow
-static -g -c -I/home/qcteam/METv4.1/include -I/h/common/gsl-
1.15/include -I/h/common/netcdf-4.1.3/include
>>> /usr/bin/g++ -m32 -o ensemble_stat ensemble_stat.cc
ensemble_stat_conf_info.o \
>>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>>> -I/home/qcteam/METv4.1/include -I/h/common/netcdf-
4.1.3/include -I/h/common/gsl-1.15/include \
>>> -L/home/qcteam/METv4.1/lib -L/h/common/netcdf-4.1.3/lib
-L/h/common/gsl-1.15/lib \
>>> -lvx_stat_out \
>>> -lvx_statistics \
>>> -lvx_shapedata \
>>> -lvx_gsl_prob \
>>> -lvx_analysis_util \
>>> -lvx_data2d_factory \
>>> -lvx_data2d_nc_met \
>>> -lvx_data2d_grib \
>>> -lvx_data2d_nc_pinterp \
>>> -lvx_data2d_nccf \
>>> -lvx_data2d \
>>> -lvx_nc_util \
>>> -lvx_grid \
>>> -lvx_config \
>>> -lvx_cal \
>>> -lvx_util \
>>> -lvx_math \
>>> -lvx_color \
>>> -lvx_log \
>>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas \
>>>
>>> /usr/bin/ld: skipping incompatible
>>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>>> -lnetcdf_c++
>>> /usr/bin/ld: cannot find -lnetcdf_c++
>>> collect2: ld returned 1 exit status
>>> make[4]: *** [ensemble_stat] Error 1
>>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>> make[3]: *** [all] Error 2
>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>>> make[2]: *** [all] Error 2
>>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>> make[1]: *** [all] Error 2
>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>> make: *** [all] Error 2
>>>
>>> -----Original Message-----
>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>> Sent: Wednesday, June 12, 2013 11:43 AM
>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>
>>> Matt,
>>>
>>> We've typically been compiling MET using NetCDF version 3 - the
specific version is netcdf-3.6.3.
>>>
>>> However, I just downloaded netcdf-4.1.3 and compiled it with the
GNU compilers (gcc and g++) using the same options you did (--disable-
netcdf-4 --disable-dap). Then I recompiled METv4.1 using netcdf-
4.1.3, and it worked fine.
>>>
>>> Your netcdf build definitely needs to have include/netcdf.hh - so
that why 4.3.0 wasn't working.
>>>
>>> Can you please do a 'make clean' followed by 'make >&
make_met.log' when compiling with netcdf-4.1.3 and then send me the
log file? What type of compilers are you using? Also, it's possible
that if the version of the compiler has changed since the time you
compiled netcdf-4.1.3 to now, that could be causing a problem.
>>>
>>> Thanks,
>>> John
>>>
>>> On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>>
>>>> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
>>>> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
>>>> Queue: met_help
>>>> Subject: NETCDF Issue
>>>> Owner: Nobody
>>>> Requestors: matthew.sittel.ctr at offutt.af.mil
>>>> Status: new
>>>> Ticket <URL:
>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830
>>>>>
>>>>
>>>>
>>>> Hi,
>>>>
>>>> Okay, now I've *really* got a problem with the compile that I
can't solve.
>>>>
>>>> Which version of NetCDF did you use to get METv4.1 to compile?
The compile I am running here at AFWA failed for version 4.1.3 (4.1.3
works great for the compile of METv4.0 though), and also fails for the
latest version of netcdf, 4.3.0. With the latter case, here's where
the compilation fails:
>>>>
>>>> *** Finished Making the vx_grid Library ***
>>>>
>>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
>>>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
>>>> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/compiled/include
>>>> In file included from nc_var_info.cc:24:
>>>> nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
>>>>
>>>> The nectdf 'include' I compiled for 4.3.0 has only one file in
it, netcdf.h, while 4.1.3 has a few more, including the netcdf.hh that
adds the inclusion of one more file from the 4.1.3 include
(netcdfcpp.h).
>>>>
>>>> The routine say netcdf compiled successfully for me, but I had to
disable a couple of things to get it to run correctly: (I ran the
following: "./configure --prefix=/home/qcteam/netcdf/compiled
--disable-netcdf-4 --disable-dap"). Is it possible one of those
disables is the reason for the missing file(s)?
>>>>
>>>> It looks as if the timing of the 4.1 release might be using an
older netcdf than 4.3.0? Rather than try every iteration of netcdf
between 4.1.3 and 4.3.0, I thought I'd ask. I wasn't going to update
netcdf but the compile of 'pb2nc' while using 4.1.3 failed, suggesting
I had to update.
>>>>
>>>> Thanks,
>>>> Matt
>>>>
>>>> // SIGNED //
>>>> Matthew C. Sittel
>>>> Associate Scientist
>>>> University Corporation for Atmospheric Research 16WS/WXN, Offutt
>>>> AFB, NE
>>>> (402) 294-3473 DSN: 271-3473
>>>>
>>>
>>>
>>
>>
>
>
------------------------------------------------
Subject: NETCDF Issue
From: matthew.sittel.ctr at offutt.af.mil
Time: Wed Jun 12 14:12:23 2013
You got it... that took care of the TC compile. Thanks again for all
your help!
Matt
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Wednesday, June 12, 2013 3:01 PM
To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
Matt,
If you don't have the HDF and HDF-EOS libraries set up, the MODIS
piece won't compile.
Regarding the problem with compiling tc_utils, I think you're missing
the "-fno-second-underscore" option in the FC_FLAGS line. It's
included in the default user_defs_gnu.mk, but for some reason it
appears not to be present in yours. Please try adding that back in,
enable the compilation of the tc_utils, and try again.
Thanks,
John
On 06/12/2013 01:26 PM, matthew.sittel.ctr at offutt.af.mil via RT wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>
> I got through that step after removing the F2C settings, as you
suggested.
>
> I got everything else to compile *if* I left out modis_regrid and
the tropical steps (i.e., set those last two options to 0 in the mk
file).
>
> Modis mentioned something about "hdf", which I think is part of the
NetCDF routine-I recall reading that in the "INSTALL" part of NetCDF.
>
> The error I got from the tropical part might be useful to you?
>
> ===
> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
>
> /usr/bin/gfortran acerr.f -Wall -Wshadow -static -ff2c -g -c
>
> /usr/bin/gfortran clip_shif.f -Wall -Wshadow -static -ff2c -g -c >
> /dev/null 2>&1
>
> /usr/bin/g++ tc_pairs_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>
> /usr/bin/g++ -o tc_pairs tc_pairs.cc acerr.o clip_shif.o
tc_pairs_conf_info.o \
> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
> -I/home/qcteam/METv4.1/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include -I/h/common/gsl-
1.15/include \
> -L/home/qcteam/METv4.1/lib -L/home/qcteam/netcdf/netcdf-
4.1.3/compiled/lib -L/h/common/gsl-1.15/lib \
> -lvx_config \
> -lvx_data2d_nc_met \
> -lvx_analysis_util \
> -lvx_tc_util \
> -lvx_nc_util \
> -lvx_grid \
> -lvx_config \
> -lvx_util \
> -lvx_math \
> -lvx_cal \
> -lvx_log \
> -lm \
> -lnetcdf_c++ -lnetcdf \
> -lgfortran
> /tmp/ccJGmQPZ.o: In function `derive_baseline_model':
> /home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs/tc_pairs.cc:1229:
undefined reference to `oclipd5_test_'
> collect2: ld returned 1 exit status
> make[4]: *** [tc_pairs] Error 1
> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
> make[3]: *** [all] Error 2
> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils'
> make[2]: *** [all] Error 2
> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
> make[1]: *** [all] Error 2
> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
> make: *** [all] Error 2
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Wednesday, June 12, 2013 1:58 PM
> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>
> Matt,
>
> I read through some previous MET-Help questions and have a
suggestion for you to try - in user_defs.mk just empty out all of the
"F2C_..." settings:
> F2C_BASE =
> F2C_INCS =
> F2C_LIBS =
> F2C_LIBNAME =
>
> Based on the log output you sent, it looks like you've specified
something in those settings. That's only necessary for the old GNU
g77, and you're using gfortran.
>
> Please let me know if that does the trick.
>
> Thanks,
> John
>
> On 06/12/2013 12:11 PM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>
>> Hey John,
>>
>> Okay, I re-compiled netcdf 4.1.3 successfully, and sourced it.
Here’s the error I get now. Since I get the same errors when
compiling with the older netcdf we already had installed and compiled
in 2011, I think we can rule out netcdf as the issue.
>>
>> =====
>>
>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>>
>> /usr/bin/g++ pb2nc_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include
>>
>> /usr/bin/gfortran numpbmsg.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran openpb.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran closepb.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran readpb.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran dumppb.f -Wall -Wshadow -static -ff2c -g -c
>> /usr/bin/g++ -o pb2nc pb2nc.cc pb2nc_conf_info.o numpbmsg.o
openpb.o closepb.o readpb.o dumppb.o \
>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>> -I/home/qcteam/METv4.1/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include -I/h/common/gsl-
1.15/include -I/h/common/BUFRLIB_v10-0-1 -I/h/common/libf2c \
>> -L/home/qcteam/METv4.1/lib -L/home/qcteam/netcdf/netcdf-
4.1.3/compiled/lib -L/h/common/gsl-1.15/lib -L/h/common/BUFRLIB_v10-0-
1 -L/h/common/libf2c \
>> -lvx_stat_out \
>> -lvx_statistics \
>> -lvx_shapedata \
>> -lvx_gsl_prob \
>> -lvx_analysis_util \
>> -lvx_data2d_grib \
>> -lvx_data2d \
>> -lvx_nc_util \
>> -lvx_grid \
>> -lvx_config \
>> -lvx_pb_util \
>> -lvx_cal \
>> -lvx_util \
>> -lvx_math \
>> -lvx_color \
>> -lvx_log \
>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas -lbufr \
>> -lf2c -lgfortran
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:70: error: \u2018::isalnum\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:71: error: \u2018::isalpha\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:72: error: \u2018::iscntrl\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:73: error: \u2018::isdigit\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:74: error: \u2018::isgraph\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:75: error: \u2018::islower\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:76: error: \u2018::isprint\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:77: error: \u2018::ispunct\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:78: error: \u2018::isspace\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:79: error: \u2018::isupper\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:80: error: \u2018::isxdigit\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:81: error: \u2018::tolower\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:82: error: \u2018::toupper\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:51: error:
\u2018_ISupper\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:52: error:
\u2018_ISlower\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:53: error:
\u2018_ISalpha\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:54: error:
\u2018_ISdigit\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:55: error:
\u2018_ISxdigit\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:56: error:
\u2018_ISspace\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:57: error:
\u2018_ISprint\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISalpha\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISdigit\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISpunct\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:59: error:
\u2018_IScntrl\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:60: error:
\u2018_ISpunct\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
\u2018_ISalpha\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
\u2018_ISdigit\u2019
>> was not declared in this scope
>> make[4]: *** [pb2nc] Error 1
>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>> make[3]: *** [all] Error 2
>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other'
>> make[2]: *** [all] Error 2
>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>> make[1]: *** [all] Error 2
>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>> make: *** [all] Error 2
>>
>> -----Original Message-----
>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>> Sent: Wednesday, June 12, 2013 12:08 PM
>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>
>> Matt,
>>
>> The answer is in the error message you sent me:
>> /usr/bin/ld: skipping incompatible
>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>> -lnetcdf_c++
>>
>> The linker isn't using the version of the NetCDF library you told
it to because it is "incompatible". In my experience, that either
means that that library was built using a different family of
compilers (e.g. PGI vs GNU) or that it was built using an earlier
version of the GNU compilers. I suspect that if you tried to rebuild
METv4.0 right now with the current version of the GNU compilers, you'd
see the same linker error.
>>
>> I'd suggest rebuilding the NetCDF library in /h/common/netcdf-4.1.3
using the current version of the GNU compilers and then try again.
>>
>> Hope that helps.
>>
>> Thanks,
>> John
>>
>> On 06/12/2013 11:00 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>
>>> OK, let me try again using the 4.1.3 we know works (at least for
v4.0). Here's where it fails during the MET v4.1 build:
>>>
>>> make[2]: Entering directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>>
>>> /usr/bin/g++ -m32 ensemble_stat_conf_info.cc -Wall -Wshadow
-static -g -c -I/home/qcteam/METv4.1/include -I/h/common/gsl-
1.15/include -I/h/common/netcdf-4.1.3/include
>>> /usr/bin/g++ -m32 -o ensemble_stat ensemble_stat.cc
ensemble_stat_conf_info.o \
>>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>>> -I/home/qcteam/METv4.1/include -I/h/common/netcdf-
4.1.3/include -I/h/common/gsl-1.15/include \
>>> -L/home/qcteam/METv4.1/lib -L/h/common/netcdf-4.1.3/lib
-L/h/common/gsl-1.15/lib \
>>> -lvx_stat_out \
>>> -lvx_statistics \
>>> -lvx_shapedata \
>>> -lvx_gsl_prob \
>>> -lvx_analysis_util \
>>> -lvx_data2d_factory \
>>> -lvx_data2d_nc_met \
>>> -lvx_data2d_grib \
>>> -lvx_data2d_nc_pinterp \
>>> -lvx_data2d_nccf \
>>> -lvx_data2d \
>>> -lvx_nc_util \
>>> -lvx_grid \
>>> -lvx_config \
>>> -lvx_cal \
>>> -lvx_util \
>>> -lvx_math \
>>> -lvx_color \
>>> -lvx_log \
>>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas \
>>>
>>> /usr/bin/ld: skipping incompatible
>>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>>> -lnetcdf_c++
>>> /usr/bin/ld: cannot find -lnetcdf_c++
>>> collect2: ld returned 1 exit status
>>> make[4]: *** [ensemble_stat] Error 1
>>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>> make[3]: *** [all] Error 2
>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>>> make[2]: *** [all] Error 2
>>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>> make[1]: *** [all] Error 2
>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>> make: *** [all] Error 2
>>>
>>> -----Original Message-----
>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>> Sent: Wednesday, June 12, 2013 11:43 AM
>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>
>>> Matt,
>>>
>>> We've typically been compiling MET using NetCDF version 3 - the
specific version is netcdf-3.6.3.
>>>
>>> However, I just downloaded netcdf-4.1.3 and compiled it with the
GNU compilers (gcc and g++) using the same options you did (--disable-
netcdf-4 --disable-dap). Then I recompiled METv4.1 using netcdf-
4.1.3, and it worked fine.
>>>
>>> Your netcdf build definitely needs to have include/netcdf.hh - so
that why 4.3.0 wasn't working.
>>>
>>> Can you please do a 'make clean' followed by 'make >&
make_met.log' when compiling with netcdf-4.1.3 and then send me the
log file? What type of compilers are you using? Also, it's possible
that if the version of the compiler has changed since the time you
compiled netcdf-4.1.3 to now, that could be causing a problem.
>>>
>>> Thanks,
>>> John
>>>
>>> On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>>
>>>> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
>>>> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
>>>> Queue: met_help
>>>> Subject: NETCDF Issue
>>>> Owner: Nobody
>>>> Requestors: matthew.sittel.ctr at offutt.af.mil
>>>> Status: new
>>>> Ticket <URL:
>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830
>>>>>
>>>>
>>>>
>>>> Hi,
>>>>
>>>> Okay, now I've *really* got a problem with the compile that I
can't solve.
>>>>
>>>> Which version of NetCDF did you use to get METv4.1 to compile?
The compile I am running here at AFWA failed for version 4.1.3 (4.1.3
works great for the compile of METv4.0 though), and also fails for the
latest version of netcdf, 4.3.0. With the latter case, here's where
the compilation fails:
>>>>
>>>> *** Finished Making the vx_grid Library ***
>>>>
>>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
>>>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
>>>> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/compiled/include
>>>> In file included from nc_var_info.cc:24:
>>>> nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
>>>>
>>>> The nectdf 'include' I compiled for 4.3.0 has only one file in
it, netcdf.h, while 4.1.3 has a few more, including the netcdf.hh that
adds the inclusion of one more file from the 4.1.3 include
(netcdfcpp.h).
>>>>
>>>> The routine say netcdf compiled successfully for me, but I had to
disable a couple of things to get it to run correctly: (I ran the
following: "./configure --prefix=/home/qcteam/netcdf/compiled
--disable-netcdf-4 --disable-dap"). Is it possible one of those
disables is the reason for the missing file(s)?
>>>>
>>>> It looks as if the timing of the 4.1 release might be using an
older netcdf than 4.3.0? Rather than try every iteration of netcdf
between 4.1.3 and 4.3.0, I thought I'd ask. I wasn't going to update
netcdf but the compile of 'pb2nc' while using 4.1.3 failed, suggesting
I had to update.
>>>>
>>>> Thanks,
>>>> Matt
>>>>
>>>> // SIGNED //
>>>> Matthew C. Sittel
>>>> Associate Scientist
>>>> University Corporation for Atmospheric Research 16WS/WXN, Offutt
>>>> AFB, NE
>>>> (402) 294-3473 DSN: 271-3473
>>>>
>>>
>>>
>>
>>
>
>
------------------------------------------------
Subject: NETCDF Issue
From: Craig, Robert J Civ USAF AFWA 16 WS/WXN
Time: Wed Jun 12 14:53:18 2013
John, now that Matt has the software compiled, I tried to run my mode
program. I get the following error:
/home/qcteam/METv4.1/bin/mode
/home/qcteam/fcsts_t001_2013061100_000.GR1
/h/data/global/WXQC/data/temp/met/cloud_obs_file_71
/home/qcteam/METv4.1/data/config/DCFModeConfig_55 -outdir
/home/qcteam/METv4.1/out/mode/ -v 10
DEBUG 1: Default Config File:
/home/qcteam/METv4.1/data/config/MODEConfig_default
DEBUG 1: Match Config File:
/home/qcteam/METv4.1/data/config/DCFModeConfig_55
DEBUG 1: Merge Config File:
/home/qcteam/METv4.1/data/config/DCFModeConfig_55
DEBUG 4: Met2dDataFileFactory::new_met_2d_data_file() -> created new
Met2dDataFile object of type "FileType_None".
ERROR :
ERROR : Trouble reading forecast file
"/home/qcteam/fcsts_t001_2013061100_000.GR1"
ERROR :
A degrib of the file header is attached (gribrecords). The file is
wide open on read permissions. I ftp the file to your site -
craig_data. I did try TCDC instead of 71 in the grib name field but
got the same error.
Thanks
Bob
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Wednesday, June 12, 2013 3:01 PM
To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
Matt,
If you don't have the HDF and HDF-EOS libraries set up, the MODIS
piece won't compile.
Regarding the problem with compiling tc_utils, I think you're missing
the "-fno-second-underscore" option in the FC_FLAGS line. It's
included in the default user_defs_gnu.mk, but for some reason it
appears not to be present in yours. Please try adding that back in,
enable the compilation of the tc_utils, and try again.
Thanks,
John
On 06/12/2013 01:26 PM, matthew.sittel.ctr at offutt.af.mil via RT wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>
> I got through that step after removing the F2C settings, as you
suggested.
>
> I got everything else to compile *if* I left out modis_regrid and
the tropical steps (i.e., set those last two options to 0 in the mk
file).
>
> Modis mentioned something about "hdf", which I think is part of the
NetCDF routine-I recall reading that in the "INSTALL" part of NetCDF.
>
> The error I got from the tropical part might be useful to you?
>
> ===
> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
>
> /usr/bin/gfortran acerr.f -Wall -Wshadow -static -ff2c -g -c
>
> /usr/bin/gfortran clip_shif.f -Wall -Wshadow -static -ff2c -g -c >
> /dev/null 2>&1
>
> /usr/bin/g++ tc_pairs_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>
> /usr/bin/g++ -o tc_pairs tc_pairs.cc acerr.o clip_shif.o
tc_pairs_conf_info.o \
> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
> -I/home/qcteam/METv4.1/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include -I/h/common/gsl-
1.15/include \
> -L/home/qcteam/METv4.1/lib -L/home/qcteam/netcdf/netcdf-
4.1.3/compiled/lib -L/h/common/gsl-1.15/lib \
> -lvx_config \
> -lvx_data2d_nc_met \
> -lvx_analysis_util \
> -lvx_tc_util \
> -lvx_nc_util \
> -lvx_grid \
> -lvx_config \
> -lvx_util \
> -lvx_math \
> -lvx_cal \
> -lvx_log \
> -lm \
> -lnetcdf_c++ -lnetcdf \
> -lgfortran
> /tmp/ccJGmQPZ.o: In function `derive_baseline_model':
> /home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs/tc_pairs.cc:1229:
undefined reference to `oclipd5_test_'
> collect2: ld returned 1 exit status
> make[4]: *** [tc_pairs] Error 1
> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
> make[3]: *** [all] Error 2
> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils'
> make[2]: *** [all] Error 2
> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
> make[1]: *** [all] Error 2
> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
> make: *** [all] Error 2
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Wednesday, June 12, 2013 1:58 PM
> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>
> Matt,
>
> I read through some previous MET-Help questions and have a
suggestion for you to try - in user_defs.mk just empty out all of the
"F2C_..." settings:
> F2C_BASE =
> F2C_INCS =
> F2C_LIBS =
> F2C_LIBNAME =
>
> Based on the log output you sent, it looks like you've specified
something in those settings. That's only necessary for the old GNU
g77, and you're using gfortran.
>
> Please let me know if that does the trick.
>
> Thanks,
> John
>
> On 06/12/2013 12:11 PM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>
>> Hey John,
>>
>> Okay, I re-compiled netcdf 4.1.3 successfully, and sourced it.
Here’s the error I get now. Since I get the same errors when
compiling with the older netcdf we already had installed and compiled
in 2011, I think we can rule out netcdf as the issue.
>>
>> =====
>>
>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>>
>> /usr/bin/g++ pb2nc_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include
>>
>> /usr/bin/gfortran numpbmsg.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran openpb.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran closepb.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran readpb.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran dumppb.f -Wall -Wshadow -static -ff2c -g -c
>> /usr/bin/g++ -o pb2nc pb2nc.cc pb2nc_conf_info.o numpbmsg.o
openpb.o closepb.o readpb.o dumppb.o \
>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>> -I/home/qcteam/METv4.1/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include -I/h/common/gsl-
1.15/include -I/h/common/BUFRLIB_v10-0-1 -I/h/common/libf2c \
>> -L/home/qcteam/METv4.1/lib -L/home/qcteam/netcdf/netcdf-
4.1.3/compiled/lib -L/h/common/gsl-1.15/lib -L/h/common/BUFRLIB_v10-0-
1 -L/h/common/libf2c \
>> -lvx_stat_out \
>> -lvx_statistics \
>> -lvx_shapedata \
>> -lvx_gsl_prob \
>> -lvx_analysis_util \
>> -lvx_data2d_grib \
>> -lvx_data2d \
>> -lvx_nc_util \
>> -lvx_grid \
>> -lvx_config \
>> -lvx_pb_util \
>> -lvx_cal \
>> -lvx_util \
>> -lvx_math \
>> -lvx_color \
>> -lvx_log \
>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas -lbufr \
>> -lf2c -lgfortran
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:70: error: \u2018::isalnum\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:71: error: \u2018::isalpha\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:72: error: \u2018::iscntrl\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:73: error: \u2018::isdigit\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:74: error: \u2018::isgraph\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:75: error: \u2018::islower\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:76: error: \u2018::isprint\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:77: error: \u2018::ispunct\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:78: error: \u2018::isspace\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:79: error: \u2018::isupper\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:80: error: \u2018::isxdigit\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:81: error: \u2018::tolower\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> c
>> ctype:82: error: \u2018::toupper\u2019 has not been declared
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:51: error:
\u2018_ISupper\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:52: error:
\u2018_ISlower\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:53: error:
\u2018_ISalpha\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:54: error:
\u2018_ISdigit\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:55: error:
\u2018_ISxdigit\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:56: error:
\u2018_ISspace\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:57: error:
\u2018_ISprint\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISalpha\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISdigit\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISpunct\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:59: error:
\u2018_IScntrl\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:60: error:
\u2018_ISpunct\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
\u2018_ISalpha\u2019
>> was not declared in this scope
>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>> x
>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
\u2018_ISdigit\u2019
>> was not declared in this scope
>> make[4]: *** [pb2nc] Error 1
>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>> make[3]: *** [all] Error 2
>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other'
>> make[2]: *** [all] Error 2
>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>> make[1]: *** [all] Error 2
>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>> make: *** [all] Error 2
>>
>> -----Original Message-----
>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>> Sent: Wednesday, June 12, 2013 12:08 PM
>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>
>> Matt,
>>
>> The answer is in the error message you sent me:
>> /usr/bin/ld: skipping incompatible
>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>> -lnetcdf_c++
>>
>> The linker isn't using the version of the NetCDF library you told
it to because it is "incompatible". In my experience, that either
means that that library was built using a different family of
compilers (e.g. PGI vs GNU) or that it was built using an earlier
version of the GNU compilers. I suspect that if you tried to rebuild
METv4.0 right now with the current version of the GNU compilers, you'd
see the same linker error.
>>
>> I'd suggest rebuilding the NetCDF library in /h/common/netcdf-4.1.3
using the current version of the GNU compilers and then try again.
>>
>> Hope that helps.
>>
>> Thanks,
>> John
>>
>> On 06/12/2013 11:00 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>
>>> OK, let me try again using the 4.1.3 we know works (at least for
v4.0). Here's where it fails during the MET v4.1 build:
>>>
>>> make[2]: Entering directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>>
>>> /usr/bin/g++ -m32 ensemble_stat_conf_info.cc -Wall -Wshadow
-static -g -c -I/home/qcteam/METv4.1/include -I/h/common/gsl-
1.15/include -I/h/common/netcdf-4.1.3/include
>>> /usr/bin/g++ -m32 -o ensemble_stat ensemble_stat.cc
ensemble_stat_conf_info.o \
>>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>>> -I/home/qcteam/METv4.1/include -I/h/common/netcdf-
4.1.3/include -I/h/common/gsl-1.15/include \
>>> -L/home/qcteam/METv4.1/lib -L/h/common/netcdf-4.1.3/lib
-L/h/common/gsl-1.15/lib \
>>> -lvx_stat_out \
>>> -lvx_statistics \
>>> -lvx_shapedata \
>>> -lvx_gsl_prob \
>>> -lvx_analysis_util \
>>> -lvx_data2d_factory \
>>> -lvx_data2d_nc_met \
>>> -lvx_data2d_grib \
>>> -lvx_data2d_nc_pinterp \
>>> -lvx_data2d_nccf \
>>> -lvx_data2d \
>>> -lvx_nc_util \
>>> -lvx_grid \
>>> -lvx_config \
>>> -lvx_cal \
>>> -lvx_util \
>>> -lvx_math \
>>> -lvx_color \
>>> -lvx_log \
>>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas \
>>>
>>> /usr/bin/ld: skipping incompatible
>>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>>> -lnetcdf_c++
>>> /usr/bin/ld: cannot find -lnetcdf_c++
>>> collect2: ld returned 1 exit status
>>> make[4]: *** [ensemble_stat] Error 1
>>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>> make[3]: *** [all] Error 2
>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>>> make[2]: *** [all] Error 2
>>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>> make[1]: *** [all] Error 2
>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>> make: *** [all] Error 2
>>>
>>> -----Original Message-----
>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>> Sent: Wednesday, June 12, 2013 11:43 AM
>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>
>>> Matt,
>>>
>>> We've typically been compiling MET using NetCDF version 3 - the
specific version is netcdf-3.6.3.
>>>
>>> However, I just downloaded netcdf-4.1.3 and compiled it with the
GNU compilers (gcc and g++) using the same options you did (--disable-
netcdf-4 --disable-dap). Then I recompiled METv4.1 using netcdf-
4.1.3, and it worked fine.
>>>
>>> Your netcdf build definitely needs to have include/netcdf.hh - so
that why 4.3.0 wasn't working.
>>>
>>> Can you please do a 'make clean' followed by 'make >&
make_met.log' when compiling with netcdf-4.1.3 and then send me the
log file? What type of compilers are you using? Also, it's possible
that if the version of the compiler has changed since the time you
compiled netcdf-4.1.3 to now, that could be causing a problem.
>>>
>>> Thanks,
>>> John
>>>
>>> On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>>
>>>> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
>>>> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
>>>> Queue: met_help
>>>> Subject: NETCDF Issue
>>>> Owner: Nobody
>>>> Requestors: matthew.sittel.ctr at offutt.af.mil
>>>> Status: new
>>>> Ticket <URL:
>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830
>>>>>
>>>>
>>>>
>>>> Hi,
>>>>
>>>> Okay, now I've *really* got a problem with the compile that I
can't solve.
>>>>
>>>> Which version of NetCDF did you use to get METv4.1 to compile?
The compile I am running here at AFWA failed for version 4.1.3 (4.1.3
works great for the compile of METv4.0 though), and also fails for the
latest version of netcdf, 4.3.0. With the latter case, here's where
the compilation fails:
>>>>
>>>> *** Finished Making the vx_grid Library ***
>>>>
>>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
>>>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
>>>> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/compiled/include
>>>> In file included from nc_var_info.cc:24:
>>>> nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
>>>>
>>>> The nectdf 'include' I compiled for 4.3.0 has only one file in
it, netcdf.h, while 4.1.3 has a few more, including the netcdf.hh that
adds the inclusion of one more file from the 4.1.3 include
(netcdfcpp.h).
>>>>
>>>> The routine say netcdf compiled successfully for me, but I had to
disable a couple of things to get it to run correctly: (I ran the
following: "./configure --prefix=/home/qcteam/netcdf/compiled
--disable-netcdf-4 --disable-dap"). Is it possible one of those
disables is the reason for the missing file(s)?
>>>>
>>>> It looks as if the timing of the 4.1 release might be using an
older netcdf than 4.3.0? Rather than try every iteration of netcdf
between 4.1.3 and 4.3.0, I thought I'd ask. I wasn't going to update
netcdf but the compile of 'pb2nc' while using 4.1.3 failed, suggesting
I had to update.
>>>>
>>>> Thanks,
>>>> Matt
>>>>
>>>> // SIGNED //
>>>> Matthew C. Sittel
>>>> Associate Scientist
>>>> University Corporation for Atmospheric Research 16WS/WXN, Offutt
>>>> AFB, NE
>>>> (402) 294-3473 DSN: 271-3473
>>>>
>>>
>>>
>>
>>
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
From: John Halley Gotway
Time: Wed Jun 12 15:28:10 2013
Bob,
I think you've uncovered a small bug in the code that figures out the
data file format to process. It isn't figuring out that the input
file is in GRIB1 format.
I'll take a look at the logic tomorrow to see what's going on. If the
input files end in ".grb" it would run fine. Or you can setup the
configuration file using:
fcst = {
field = {
name = "TCDC";
level = "L0";
};
raw_thresh = >=0.0;
conv_radius = 60.0/grid_res; // in grid squares
conv_thresh = >=81.0;
vld_thresh = 0.5;
area_thresh = >=20.0;
inten_perc_value = 100;
inten_perc_thresh = >=0.0;
merge_thresh = >=1.25;
merge_flag = THRESH;
file_type = GRIB1;
};
Here I've added "file_type = GRIB1;". That overrides the file type
guesser and explicitly states how to read the file.
Note also that I've set "name = "TCDC";".
Hopefully that'll get you going.
Thanks,
John
On 06/12/2013 02:53 PM, Craig, Robert J Civ USAF AFWA 16 WS/WXN via RT
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>
> John, now that Matt has the software compiled, I tried to run my
mode program. I get the following error:
>
> /home/qcteam/METv4.1/bin/mode
/home/qcteam/fcsts_t001_2013061100_000.GR1
/h/data/global/WXQC/data/temp/met/cloud_obs_file_71
/home/qcteam/METv4.1/data/config/DCFModeConfig_55 -outdir
/home/qcteam/METv4.1/out/mode/ -v 10
> DEBUG 1: Default Config File:
/home/qcteam/METv4.1/data/config/MODEConfig_default
> DEBUG 1: Match Config File:
/home/qcteam/METv4.1/data/config/DCFModeConfig_55
> DEBUG 1: Merge Config File:
/home/qcteam/METv4.1/data/config/DCFModeConfig_55
> DEBUG 4: Met2dDataFileFactory::new_met_2d_data_file() -> created new
Met2dDataFile object of type "FileType_None".
> ERROR :
> ERROR : Trouble reading forecast file
"/home/qcteam/fcsts_t001_2013061100_000.GR1"
> ERROR :
>
> A degrib of the file header is attached (gribrecords). The file is
wide open on read permissions. I ftp the file to your site -
craig_data. I did try TCDC instead of 71 in the grib name field but
got the same error.
>
> Thanks
>
> Bob
>
>
>
>
>
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Wednesday, June 12, 2013 3:01 PM
> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>
> Matt,
>
> If you don't have the HDF and HDF-EOS libraries set up, the MODIS
piece won't compile.
>
> Regarding the problem with compiling tc_utils, I think you're
missing the "-fno-second-underscore" option in the FC_FLAGS line.
It's included in the default user_defs_gnu.mk, but for some reason it
appears not to be present in yours. Please try adding that back in,
enable the compilation of the tc_utils, and try again.
>
> Thanks,
> John
>
> On 06/12/2013 01:26 PM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>
>> I got through that step after removing the F2C settings, as you
suggested.
>>
>> I got everything else to compile *if* I left out modis_regrid and
the tropical steps (i.e., set those last two options to 0 in the mk
file).
>>
>> Modis mentioned something about "hdf", which I think is part of the
NetCDF routine-I recall reading that in the "INSTALL" part of NetCDF.
>>
>> The error I got from the tropical part might be useful to you?
>>
>> ===
>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
>>
>> /usr/bin/gfortran acerr.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran clip_shif.f -Wall -Wshadow -static -ff2c -g -c >
>> /dev/null 2>&1
>>
>> /usr/bin/g++ tc_pairs_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>>
>> /usr/bin/g++ -o tc_pairs tc_pairs.cc acerr.o clip_shif.o
tc_pairs_conf_info.o \
>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>> -I/home/qcteam/METv4.1/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include -I/h/common/gsl-
1.15/include \
>> -L/home/qcteam/METv4.1/lib -L/home/qcteam/netcdf/netcdf-
4.1.3/compiled/lib -L/h/common/gsl-1.15/lib \
>> -lvx_config \
>> -lvx_data2d_nc_met \
>> -lvx_analysis_util \
>> -lvx_tc_util \
>> -lvx_nc_util \
>> -lvx_grid \
>> -lvx_config \
>> -lvx_util \
>> -lvx_math \
>> -lvx_cal \
>> -lvx_log \
>> -lm \
>> -lnetcdf_c++ -lnetcdf \
>> -lgfortran
>> /tmp/ccJGmQPZ.o: In function `derive_baseline_model':
>> /home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs/tc_pairs.cc:1229:
undefined reference to `oclipd5_test_'
>> collect2: ld returned 1 exit status
>> make[4]: *** [tc_pairs] Error 1
>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
>> make[3]: *** [all] Error 2
>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils'
>> make[2]: *** [all] Error 2
>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>> make[1]: *** [all] Error 2
>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>> make: *** [all] Error 2
>>
>> -----Original Message-----
>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>> Sent: Wednesday, June 12, 2013 1:58 PM
>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>
>> Matt,
>>
>> I read through some previous MET-Help questions and have a
suggestion for you to try - in user_defs.mk just empty out all of the
"F2C_..." settings:
>> F2C_BASE =
>> F2C_INCS =
>> F2C_LIBS =
>> F2C_LIBNAME =
>>
>> Based on the log output you sent, it looks like you've specified
something in those settings. That's only necessary for the old GNU
g77, and you're using gfortran.
>>
>> Please let me know if that does the trick.
>>
>> Thanks,
>> John
>>
>> On 06/12/2013 12:11 PM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>
>>> Hey John,
>>>
>>> Okay, I re-compiled netcdf 4.1.3 successfully, and sourced it.
Here’s the error I get now. Since I get the same errors when
compiling with the older netcdf we already had installed and compiled
in 2011, I think we can rule out netcdf as the issue.
>>>
>>> =====
>>>
>>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>>>
>>> /usr/bin/g++ pb2nc_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include
>>>
>>> /usr/bin/gfortran numpbmsg.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran openpb.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran closepb.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran readpb.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran dumppb.f -Wall -Wshadow -static -ff2c -g -c
>>> /usr/bin/g++ -o pb2nc pb2nc.cc pb2nc_conf_info.o numpbmsg.o
openpb.o closepb.o readpb.o dumppb.o \
>>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>>> -I/home/qcteam/METv4.1/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include -I/h/common/gsl-
1.15/include -I/h/common/BUFRLIB_v10-0-1 -I/h/common/libf2c \
>>> -L/home/qcteam/METv4.1/lib
-L/home/qcteam/netcdf/netcdf-4.1.3/compiled/lib -L/h/common/gsl-
1.15/lib -L/h/common/BUFRLIB_v10-0-1 -L/h/common/libf2c \
>>> -lvx_stat_out \
>>> -lvx_statistics \
>>> -lvx_shapedata \
>>> -lvx_gsl_prob \
>>> -lvx_analysis_util \
>>> -lvx_data2d_grib \
>>> -lvx_data2d \
>>> -lvx_nc_util \
>>> -lvx_grid \
>>> -lvx_config \
>>> -lvx_pb_util \
>>> -lvx_cal \
>>> -lvx_util \
>>> -lvx_math \
>>> -lvx_color \
>>> -lvx_log \
>>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas -lbufr \
>>> -lf2c -lgfortran
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:70: error: \u2018::isalnum\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:71: error: \u2018::isalpha\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:72: error: \u2018::iscntrl\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:73: error: \u2018::isdigit\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:74: error: \u2018::isgraph\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:75: error: \u2018::islower\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:76: error: \u2018::isprint\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:77: error: \u2018::ispunct\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:78: error: \u2018::isspace\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:79: error: \u2018::isupper\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:80: error: \u2018::isxdigit\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:81: error: \u2018::tolower\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:82: error: \u2018::toupper\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:51: error:
\u2018_ISupper\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:52: error:
\u2018_ISlower\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:53: error:
\u2018_ISalpha\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:54: error:
\u2018_ISdigit\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:55: error:
\u2018_ISxdigit\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:56: error:
\u2018_ISspace\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:57: error:
\u2018_ISprint\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISalpha\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISdigit\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISpunct\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:59: error:
\u2018_IScntrl\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:60: error:
\u2018_ISpunct\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
\u2018_ISalpha\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
\u2018_ISdigit\u2019
>>> was not declared in this scope
>>> make[4]: *** [pb2nc] Error 1
>>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>>> make[3]: *** [all] Error 2
>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other'
>>> make[2]: *** [all] Error 2
>>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>> make[1]: *** [all] Error 2
>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>> make: *** [all] Error 2
>>>
>>> -----Original Message-----
>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>> Sent: Wednesday, June 12, 2013 12:08 PM
>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>
>>> Matt,
>>>
>>> The answer is in the error message you sent me:
>>> /usr/bin/ld: skipping incompatible
>>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>>> -lnetcdf_c++
>>>
>>> The linker isn't using the version of the NetCDF library you told
it to because it is "incompatible". In my experience, that either
means that that library was built using a different family of
compilers (e.g. PGI vs GNU) or that it was built using an earlier
version of the GNU compilers. I suspect that if you tried to rebuild
METv4.0 right now with the current version of the GNU compilers, you'd
see the same linker error.
>>>
>>> I'd suggest rebuilding the NetCDF library in /h/common/netcdf-
4.1.3 using the current version of the GNU compilers and then try
again.
>>>
>>> Hope that helps.
>>>
>>> Thanks,
>>> John
>>>
>>> On 06/12/2013 11:00 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>>
>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>>
>>>> OK, let me try again using the 4.1.3 we know works (at least for
v4.0). Here's where it fails during the MET v4.1 build:
>>>>
>>>> make[2]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools'
>>>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>>>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>>>
>>>> /usr/bin/g++ -m32 ensemble_stat_conf_info.cc -Wall -Wshadow
-static -g -c -I/home/qcteam/METv4.1/include -I/h/common/gsl-
1.15/include -I/h/common/netcdf-4.1.3/include
>>>> /usr/bin/g++ -m32 -o ensemble_stat ensemble_stat.cc
ensemble_stat_conf_info.o \
>>>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>>>> -I/home/qcteam/METv4.1/include -I/h/common/netcdf-
4.1.3/include -I/h/common/gsl-1.15/include \
>>>> -L/home/qcteam/METv4.1/lib -L/h/common/netcdf-
4.1.3/lib -L/h/common/gsl-1.15/lib \
>>>> -lvx_stat_out \
>>>> -lvx_statistics \
>>>> -lvx_shapedata \
>>>> -lvx_gsl_prob \
>>>> -lvx_analysis_util \
>>>> -lvx_data2d_factory \
>>>> -lvx_data2d_nc_met \
>>>> -lvx_data2d_grib \
>>>> -lvx_data2d_nc_pinterp \
>>>> -lvx_data2d_nccf \
>>>> -lvx_data2d \
>>>> -lvx_nc_util \
>>>> -lvx_grid \
>>>> -lvx_config \
>>>> -lvx_cal \
>>>> -lvx_util \
>>>> -lvx_math \
>>>> -lvx_color \
>>>> -lvx_log \
>>>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas \
>>>>
>>>> /usr/bin/ld: skipping incompatible
>>>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>>>> -lnetcdf_c++
>>>> /usr/bin/ld: cannot find -lnetcdf_c++
>>>> collect2: ld returned 1 exit status
>>>> make[4]: *** [ensemble_stat] Error 1
>>>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>>> make[3]: *** [all] Error 2
>>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>>>> make[2]: *** [all] Error 2
>>>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>>> make[1]: *** [all] Error 2
>>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>>> make: *** [all] Error 2
>>>>
>>>> -----Original Message-----
>>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>>> Sent: Wednesday, June 12, 2013 11:43 AM
>>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>>
>>>> Matt,
>>>>
>>>> We've typically been compiling MET using NetCDF version 3 - the
specific version is netcdf-3.6.3.
>>>>
>>>> However, I just downloaded netcdf-4.1.3 and compiled it with the
GNU compilers (gcc and g++) using the same options you did (--disable-
netcdf-4 --disable-dap). Then I recompiled METv4.1 using netcdf-
4.1.3, and it worked fine.
>>>>
>>>> Your netcdf build definitely needs to have include/netcdf.hh - so
that why 4.3.0 wasn't working.
>>>>
>>>> Can you please do a 'make clean' followed by 'make >&
make_met.log' when compiling with netcdf-4.1.3 and then send me the
log file? What type of compilers are you using? Also, it's possible
that if the version of the compiler has changed since the time you
compiled netcdf-4.1.3 to now, that could be causing a problem.
>>>>
>>>> Thanks,
>>>> John
>>>>
>>>> On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>>>
>>>>> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
>>>>> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
>>>>> Queue: met_help
>>>>> Subject: NETCDF Issue
>>>>> Owner: Nobody
>>>>> Requestors: matthew.sittel.ctr at offutt.af.mil
>>>>> Status: new
>>>>> Ticket <URL:
>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830
>>>>>>
>>>>>
>>>>>
>>>>> Hi,
>>>>>
>>>>> Okay, now I've *really* got a problem with the compile that I
can't solve.
>>>>>
>>>>> Which version of NetCDF did you use to get METv4.1 to compile?
The compile I am running here at AFWA failed for version 4.1.3 (4.1.3
works great for the compile of METv4.0 though), and also fails for the
latest version of netcdf, 4.3.0. With the latter case, here's where
the compilation fails:
>>>>>
>>>>> *** Finished Making the vx_grid Library ***
>>>>>
>>>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
>>>>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
>>>>> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/compiled/include
>>>>> In file included from nc_var_info.cc:24:
>>>>> nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
>>>>>
>>>>> The nectdf 'include' I compiled for 4.3.0 has only one file in
it, netcdf.h, while 4.1.3 has a few more, including the netcdf.hh that
adds the inclusion of one more file from the 4.1.3 include
(netcdfcpp.h).
>>>>>
>>>>> The routine say netcdf compiled successfully for me, but I had
to disable a couple of things to get it to run correctly: (I ran the
following: "./configure --prefix=/home/qcteam/netcdf/compiled
--disable-netcdf-4 --disable-dap"). Is it possible one of those
disables is the reason for the missing file(s)?
>>>>>
>>>>> It looks as if the timing of the 4.1 release might be using an
older netcdf than 4.3.0? Rather than try every iteration of netcdf
between 4.1.3 and 4.3.0, I thought I'd ask. I wasn't going to update
netcdf but the compile of 'pb2nc' while using 4.1.3 failed, suggesting
I had to update.
>>>>>
>>>>> Thanks,
>>>>> Matt
>>>>>
>>>>> // SIGNED //
>>>>> Matthew C. Sittel
>>>>> Associate Scientist
>>>>> University Corporation for Atmospheric Research 16WS/WXN, Offutt
>>>>> AFB, NE
>>>>> (402) 294-3473 DSN: 271-3473
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
From: John Halley Gotway
Time: Wed Jun 12 15:39:29 2013
Bob,
I see why the code is not recognizing this as a GRIB file. We expect
that the first 4 characters in the file are "GRIB". That's called the
"magic cookie" to indicate the file type.
Here's what they are in the file you sent me:
od -c fcsts_t001_2013061100_000.GR1 | more
0000000 ¼ x 003 \0 G R I B 003 x º 001 \0 \0 034
203
So GRIB is in there, but just not at the beginning. I'm asking around
to find out if the problem is in your GRIB file, or in our logic when
reading it.
As I said, you can get around this either by changing the file's
suffix to be ".grb" or using the "file_type = GRIB1;" option in the
config file.
Thanks,
John
On 06/12/2013 02:53 PM, Craig, Robert J Civ USAF AFWA 16 WS/WXN via RT
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>
> John, now that Matt has the software compiled, I tried to run my
mode program. I get the following error:
>
> /home/qcteam/METv4.1/bin/mode
/home/qcteam/fcsts_t001_2013061100_000.GR1
/h/data/global/WXQC/data/temp/met/cloud_obs_file_71
/home/qcteam/METv4.1/data/config/DCFModeConfig_55 -outdir
/home/qcteam/METv4.1/out/mode/ -v 10
> DEBUG 1: Default Config File:
/home/qcteam/METv4.1/data/config/MODEConfig_default
> DEBUG 1: Match Config File:
/home/qcteam/METv4.1/data/config/DCFModeConfig_55
> DEBUG 1: Merge Config File:
/home/qcteam/METv4.1/data/config/DCFModeConfig_55
> DEBUG 4: Met2dDataFileFactory::new_met_2d_data_file() -> created new
Met2dDataFile object of type "FileType_None".
> ERROR :
> ERROR : Trouble reading forecast file
"/home/qcteam/fcsts_t001_2013061100_000.GR1"
> ERROR :
>
> A degrib of the file header is attached (gribrecords). The file is
wide open on read permissions. I ftp the file to your site -
craig_data. I did try TCDC instead of 71 in the grib name field but
got the same error.
>
> Thanks
>
> Bob
>
>
>
>
>
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Wednesday, June 12, 2013 3:01 PM
> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>
> Matt,
>
> If you don't have the HDF and HDF-EOS libraries set up, the MODIS
piece won't compile.
>
> Regarding the problem with compiling tc_utils, I think you're
missing the "-fno-second-underscore" option in the FC_FLAGS line.
It's included in the default user_defs_gnu.mk, but for some reason it
appears not to be present in yours. Please try adding that back in,
enable the compilation of the tc_utils, and try again.
>
> Thanks,
> John
>
> On 06/12/2013 01:26 PM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>
>> I got through that step after removing the F2C settings, as you
suggested.
>>
>> I got everything else to compile *if* I left out modis_regrid and
the tropical steps (i.e., set those last two options to 0 in the mk
file).
>>
>> Modis mentioned something about "hdf", which I think is part of the
NetCDF routine-I recall reading that in the "INSTALL" part of NetCDF.
>>
>> The error I got from the tropical part might be useful to you?
>>
>> ===
>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
>>
>> /usr/bin/gfortran acerr.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran clip_shif.f -Wall -Wshadow -static -ff2c -g -c >
>> /dev/null 2>&1
>>
>> /usr/bin/g++ tc_pairs_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>>
>> /usr/bin/g++ -o tc_pairs tc_pairs.cc acerr.o clip_shif.o
tc_pairs_conf_info.o \
>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>> -I/home/qcteam/METv4.1/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include -I/h/common/gsl-
1.15/include \
>> -L/home/qcteam/METv4.1/lib -L/home/qcteam/netcdf/netcdf-
4.1.3/compiled/lib -L/h/common/gsl-1.15/lib \
>> -lvx_config \
>> -lvx_data2d_nc_met \
>> -lvx_analysis_util \
>> -lvx_tc_util \
>> -lvx_nc_util \
>> -lvx_grid \
>> -lvx_config \
>> -lvx_util \
>> -lvx_math \
>> -lvx_cal \
>> -lvx_log \
>> -lm \
>> -lnetcdf_c++ -lnetcdf \
>> -lgfortran
>> /tmp/ccJGmQPZ.o: In function `derive_baseline_model':
>> /home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs/tc_pairs.cc:1229:
undefined reference to `oclipd5_test_'
>> collect2: ld returned 1 exit status
>> make[4]: *** [tc_pairs] Error 1
>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
>> make[3]: *** [all] Error 2
>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/tc_utils'
>> make[2]: *** [all] Error 2
>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>> make[1]: *** [all] Error 2
>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>> make: *** [all] Error 2
>>
>> -----Original Message-----
>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>> Sent: Wednesday, June 12, 2013 1:58 PM
>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>
>> Matt,
>>
>> I read through some previous MET-Help questions and have a
suggestion for you to try - in user_defs.mk just empty out all of the
"F2C_..." settings:
>> F2C_BASE =
>> F2C_INCS =
>> F2C_LIBS =
>> F2C_LIBNAME =
>>
>> Based on the log output you sent, it looks like you've specified
something in those settings. That's only necessary for the old GNU
g77, and you're using gfortran.
>>
>> Please let me know if that does the trick.
>>
>> Thanks,
>> John
>>
>> On 06/12/2013 12:11 PM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>
>>> Hey John,
>>>
>>> Okay, I re-compiled netcdf 4.1.3 successfully, and sourced it.
Here’s the error I get now. Since I get the same errors when
compiling with the older netcdf we already had installed and compiled
in 2011, I think we can rule out netcdf as the issue.
>>>
>>> =====
>>>
>>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>>>
>>> /usr/bin/g++ pb2nc_conf_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include
>>>
>>> /usr/bin/gfortran numpbmsg.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran openpb.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran closepb.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran readpb.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran dumppb.f -Wall -Wshadow -static -ff2c -g -c
>>> /usr/bin/g++ -o pb2nc pb2nc.cc pb2nc_conf_info.o numpbmsg.o
openpb.o closepb.o readpb.o dumppb.o \
>>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>>> -I/home/qcteam/METv4.1/include
-I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include -I/h/common/gsl-
1.15/include -I/h/common/BUFRLIB_v10-0-1 -I/h/common/libf2c \
>>> -L/home/qcteam/METv4.1/lib
-L/home/qcteam/netcdf/netcdf-4.1.3/compiled/lib -L/h/common/gsl-
1.15/lib -L/h/common/BUFRLIB_v10-0-1 -L/h/common/libf2c \
>>> -lvx_stat_out \
>>> -lvx_statistics \
>>> -lvx_shapedata \
>>> -lvx_gsl_prob \
>>> -lvx_analysis_util \
>>> -lvx_data2d_grib \
>>> -lvx_data2d \
>>> -lvx_nc_util \
>>> -lvx_grid \
>>> -lvx_config \
>>> -lvx_pb_util \
>>> -lvx_cal \
>>> -lvx_util \
>>> -lvx_math \
>>> -lvx_color \
>>> -lvx_log \
>>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas -lbufr \
>>> -lf2c -lgfortran
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:70: error: \u2018::isalnum\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:71: error: \u2018::isalpha\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:72: error: \u2018::iscntrl\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:73: error: \u2018::isdigit\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:74: error: \u2018::isgraph\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:75: error: \u2018::islower\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:76: error: \u2018::isprint\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:77: error: \u2018::ispunct\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:78: error: \u2018::isspace\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:79: error: \u2018::isupper\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:80: error: \u2018::isxdigit\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:81: error: \u2018::tolower\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:82: error: \u2018::toupper\u2019 has not been declared
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:51: error:
\u2018_ISupper\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:52: error:
\u2018_ISlower\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:53: error:
\u2018_ISalpha\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:54: error:
\u2018_ISdigit\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:55: error:
\u2018_ISxdigit\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:56: error:
\u2018_ISspace\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:57: error:
\u2018_ISprint\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISalpha\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISdigit\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
\u2018_ISpunct\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:59: error:
\u2018_IScntrl\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:60: error:
\u2018_ISpunct\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
\u2018_ISalpha\u2019
>>> was not declared in this scope
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
\u2018_ISdigit\u2019
>>> was not declared in this scope
>>> make[4]: *** [pb2nc] Error 1
>>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>>> make[3]: *** [all] Error 2
>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/other'
>>> make[2]: *** [all] Error 2
>>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>> make[1]: *** [all] Error 2
>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>> make: *** [all] Error 2
>>>
>>> -----Original Message-----
>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>> Sent: Wednesday, June 12, 2013 12:08 PM
>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>
>>> Matt,
>>>
>>> The answer is in the error message you sent me:
>>> /usr/bin/ld: skipping incompatible
>>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>>> -lnetcdf_c++
>>>
>>> The linker isn't using the version of the NetCDF library you told
it to because it is "incompatible". In my experience, that either
means that that library was built using a different family of
compilers (e.g. PGI vs GNU) or that it was built using an earlier
version of the GNU compilers. I suspect that if you tried to rebuild
METv4.0 right now with the current version of the GNU compilers, you'd
see the same linker error.
>>>
>>> I'd suggest rebuilding the NetCDF library in /h/common/netcdf-
4.1.3 using the current version of the GNU compilers and then try
again.
>>>
>>> Hope that helps.
>>>
>>> Thanks,
>>> John
>>>
>>> On 06/12/2013 11:00 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>>
>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>>
>>>> OK, let me try again using the 4.1.3 we know works (at least for
v4.0). Here's where it fails during the MET v4.1 build:
>>>>
>>>> make[2]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools'
>>>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>>>> make[4]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>>>
>>>> /usr/bin/g++ -m32 ensemble_stat_conf_info.cc -Wall -Wshadow
-static -g -c -I/home/qcteam/METv4.1/include -I/h/common/gsl-
1.15/include -I/h/common/netcdf-4.1.3/include
>>>> /usr/bin/g++ -m32 -o ensemble_stat ensemble_stat.cc
ensemble_stat_conf_info.o \
>>>> -Wall -Wshadow -static -g
-DMET_BASE=\"/home/qcteam/METv4.1\" \
>>>> -I/home/qcteam/METv4.1/include -I/h/common/netcdf-
4.1.3/include -I/h/common/gsl-1.15/include \
>>>> -L/home/qcteam/METv4.1/lib -L/h/common/netcdf-
4.1.3/lib -L/h/common/gsl-1.15/lib \
>>>> -lvx_stat_out \
>>>> -lvx_statistics \
>>>> -lvx_shapedata \
>>>> -lvx_gsl_prob \
>>>> -lvx_analysis_util \
>>>> -lvx_data2d_factory \
>>>> -lvx_data2d_nc_met \
>>>> -lvx_data2d_grib \
>>>> -lvx_data2d_nc_pinterp \
>>>> -lvx_data2d_nccf \
>>>> -lvx_data2d \
>>>> -lvx_nc_util \
>>>> -lvx_grid \
>>>> -lvx_config \
>>>> -lvx_cal \
>>>> -lvx_util \
>>>> -lvx_math \
>>>> -lvx_color \
>>>> -lvx_log \
>>>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas \
>>>>
>>>> /usr/bin/ld: skipping incompatible
>>>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>>>> -lnetcdf_c++
>>>> /usr/bin/ld: cannot find -lnetcdf_c++
>>>> collect2: ld returned 1 exit status
>>>> make[4]: *** [ensemble_stat] Error 1
>>>> make[4]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>>> make[3]: *** [all] Error 2
>>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools/core'
>>>> make[2]: *** [all] Error 2
>>>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>>> make[1]: *** [all] Error 2
>>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>>> make: *** [all] Error 2
>>>>
>>>> -----Original Message-----
>>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>>> Sent: Wednesday, June 12, 2013 11:43 AM
>>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>>
>>>> Matt,
>>>>
>>>> We've typically been compiling MET using NetCDF version 3 - the
specific version is netcdf-3.6.3.
>>>>
>>>> However, I just downloaded netcdf-4.1.3 and compiled it with the
GNU compilers (gcc and g++) using the same options you did (--disable-
netcdf-4 --disable-dap). Then I recompiled METv4.1 using netcdf-
4.1.3, and it worked fine.
>>>>
>>>> Your netcdf build definitely needs to have include/netcdf.hh - so
that why 4.3.0 wasn't working.
>>>>
>>>> Can you please do a 'make clean' followed by 'make >&
make_met.log' when compiling with netcdf-4.1.3 and then send me the
log file? What type of compilers are you using? Also, it's possible
that if the version of the compiler has changed since the time you
compiled netcdf-4.1.3 to now, that could be causing a problem.
>>>>
>>>> Thanks,
>>>> John
>>>>
>>>> On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT
wrote:
>>>>>
>>>>> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
>>>>> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
>>>>> Queue: met_help
>>>>> Subject: NETCDF Issue
>>>>> Owner: Nobody
>>>>> Requestors: matthew.sittel.ctr at offutt.af.mil
>>>>> Status: new
>>>>> Ticket <URL:
>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830
>>>>>>
>>>>>
>>>>>
>>>>> Hi,
>>>>>
>>>>> Okay, now I've *really* got a problem with the compile that I
can't solve.
>>>>>
>>>>> Which version of NetCDF did you use to get METv4.1 to compile?
The compile I am running here at AFWA failed for version 4.1.3 (4.1.3
works great for the compile of METv4.0 though), and also fails for the
latest version of netcdf, 4.3.0. With the latter case, here's where
the compilation fails:
>>>>>
>>>>> *** Finished Making the vx_grid Library ***
>>>>>
>>>>> make[3]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
>>>>> make[3]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
>>>>> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
-I/home/qcteam/METv4.1/include -I/home/qcteam/netcdf/compiled/include
>>>>> In file included from nc_var_info.cc:24:
>>>>> nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
>>>>>
>>>>> The nectdf 'include' I compiled for 4.3.0 has only one file in
it, netcdf.h, while 4.1.3 has a few more, including the netcdf.hh that
adds the inclusion of one more file from the 4.1.3 include
(netcdfcpp.h).
>>>>>
>>>>> The routine say netcdf compiled successfully for me, but I had
to disable a couple of things to get it to run correctly: (I ran the
following: "./configure --prefix=/home/qcteam/netcdf/compiled
--disable-netcdf-4 --disable-dap"). Is it possible one of those
disables is the reason for the missing file(s)?
>>>>>
>>>>> It looks as if the timing of the 4.1 release might be using an
older netcdf than 4.3.0? Rather than try every iteration of netcdf
between 4.1.3 and 4.3.0, I thought I'd ask. I wasn't going to update
netcdf but the compile of 'pb2nc' while using 4.1.3 failed, suggesting
I had to update.
>>>>>
>>>>> Thanks,
>>>>> Matt
>>>>>
>>>>> // SIGNED //
>>>>> Matthew C. Sittel
>>>>> Associate Scientist
>>>>> University Corporation for Atmospheric Research 16WS/WXN, Offutt
>>>>> AFB, NE
>>>>> (402) 294-3473 DSN: 271-3473
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #61830] NETCDF Issue
From: John Halley Gotway
Time: Thu Jun 13 13:34:25 2013
Bob,
I've posted a bugfix for METv4.1 here:
http://www.dtcenter.org/met/users/support/known_issues/METv4.1/index.php
The change should allow you to use that GRIB file you sent me without
having to change it's suffix or use the file_type option in the config
file.
Please try applying the patch and let me know that they work for you.
Thanks,
John
On 2013-06-12 14:53, Craig, Robert J Civ USAF AFWA 16 WS/WXN via RT
wrote:
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>
> John, now that Matt has the software compiled, I tried to run my
mode
> program. I get the following error:
>
> /home/qcteam/METv4.1/bin/mode
> /home/qcteam/fcsts_t001_2013061100_000.GR1
> /h/data/global/WXQC/data/temp/met/cloud_obs_file_71
> /home/qcteam/METv4.1/data/config/DCFModeConfig_55 -outdir
> /home/qcteam/METv4.1/out/mode/ -v 10
> DEBUG 1: Default Config File:
> /home/qcteam/METv4.1/data/config/MODEConfig_default
> DEBUG 1: Match Config File:
> /home/qcteam/METv4.1/data/config/DCFModeConfig_55
> DEBUG 1: Merge Config File:
> /home/qcteam/METv4.1/data/config/DCFModeConfig_55
> DEBUG 4: Met2dDataFileFactory::new_met_2d_data_file() -> created new
> Met2dDataFile object of type "FileType_None".
> ERROR :
> ERROR : Trouble reading forecast file
> "/home/qcteam/fcsts_t001_2013061100_000.GR1"
> ERROR :
>
> A degrib of the file header is attached (gribrecords). The file is
> wide open on read permissions. I ftp the file to your site -
> craig_data. I did try TCDC instead of 71 in the grib name field but
> got the same error.
>
> Thanks
>
> Bob
>
>
>
>
>
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Wednesday, June 12, 2013 3:01 PM
> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>
> Matt,
>
> If you don't have the HDF and HDF-EOS libraries set up, the MODIS
> piece won't compile.
>
> Regarding the problem with compiling tc_utils, I think you're
missing
> the "-fno-second-underscore" option in the FC_FLAGS line. It's
> included in the default user_defs_gnu.mk, but for some reason it
> appears not to be present in yours. Please try adding that back in,
> enable the compilation of the tc_utils, and try again.
>
> Thanks,
> John
>
> On 06/12/2013 01:26 PM, matthew.sittel.ctr at offutt.af.mil via RT
> wrote:
>>
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>
>> I got through that step after removing the F2C settings, as you
>> suggested.
>>
>> I got everything else to compile *if* I left out modis_regrid and
>> the tropical steps (i.e., set those last two options to 0 in the mk
>> file).
>>
>> Modis mentioned something about "hdf", which I think is part of the
>> NetCDF routine-I recall reading that in the "INSTALL" part of
NetCDF.
>>
>> The error I got from the tropical part might be useful to you?
>>
>> ===
>> make[4]: Entering directory
>> `/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
>>
>> /usr/bin/gfortran acerr.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran clip_shif.f -Wall -Wshadow -static -ff2c -g -c >
>> /dev/null 2>&1
>>
>> /usr/bin/g++ tc_pairs_conf_info.cc -Wall -Wshadow -static -g -c
>> -I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
>> -I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>>
>> /usr/bin/g++ -o tc_pairs tc_pairs.cc acerr.o clip_shif.o
>> tc_pairs_conf_info.o \
>> -Wall -Wshadow -static -g
>> -DMET_BASE=\"/home/qcteam/METv4.1\" \
>> -I/home/qcteam/METv4.1/include
>> -I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>> -I/h/common/gsl-1.15/include \
>> -L/home/qcteam/METv4.1/lib
>> -L/home/qcteam/netcdf/netcdf-4.1.3/compiled/lib
>> -L/h/common/gsl-1.15/lib \
>> -lvx_config \
>> -lvx_data2d_nc_met \
>> -lvx_analysis_util \
>> -lvx_tc_util \
>> -lvx_nc_util \
>> -lvx_grid \
>> -lvx_config \
>> -lvx_util \
>> -lvx_math \
>> -lvx_cal \
>> -lvx_log \
>> -lm \
>> -lnetcdf_c++ -lnetcdf \
>> -lgfortran
>> /tmp/ccJGmQPZ.o: In function `derive_baseline_model':
>> /home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs/tc_pairs.cc:1229:
>> undefined reference to `oclipd5_test_'
>> collect2: ld returned 1 exit status
>> make[4]: *** [tc_pairs] Error 1
>> make[4]: Leaving directory
>> `/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
>> make[3]: *** [all] Error 2
>> make[3]: Leaving directory
>> `/ibrix/home/qcteam/METv4.1/src/tools/tc_utils'
>> make[2]: *** [all] Error 2
>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>> make[1]: *** [all] Error 2
>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>> make: *** [all] Error 2
>>
>> -----Original Message-----
>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>> Sent: Wednesday, June 12, 2013 1:58 PM
>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>
>> Matt,
>>
>> I read through some previous MET-Help questions and have a
>> suggestion for you to try - in user_defs.mk just empty out all of
the
>> "F2C_..." settings:
>> F2C_BASE =
>> F2C_INCS =
>> F2C_LIBS =
>> F2C_LIBNAME =
>>
>> Based on the log output you sent, it looks like you've specified
>> something in those settings. That's only necessary for the old GNU
>> g77, and you're using gfortran.
>>
>> Please let me know if that does the trick.
>>
>> Thanks,
>> John
>>
>> On 06/12/2013 12:11 PM, matthew.sittel.ctr at offutt.af.mil via RT
>> wrote:
>>>
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>
>>> Hey John,
>>>
>>> Okay, I re-compiled netcdf 4.1.3 successfully, and sourced it.
>>> Here’s the error I get now. Since I get the same errors when
>>> compiling with the older netcdf we already had installed and
compiled
>>> in 2011, I think we can rule out netcdf as the issue.
>>>
>>> =====
>>>
>>> make[4]: Entering directory
>>> `/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>>>
>>> /usr/bin/g++ pb2nc_conf_info.cc -Wall -Wshadow -static -g -c
>>> -I/home/qcteam/METv4.1/include
>>>
>>> /usr/bin/gfortran numpbmsg.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran openpb.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran closepb.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran readpb.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran dumppb.f -Wall -Wshadow -static -ff2c -g -c
>>> /usr/bin/g++ -o pb2nc pb2nc.cc pb2nc_conf_info.o numpbmsg.o
>>> openpb.o closepb.o readpb.o dumppb.o \
>>> -Wall -Wshadow -static -g
>>> -DMET_BASE=\"/home/qcteam/METv4.1\" \
>>> -I/home/qcteam/METv4.1/include
>>> -I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>>> -I/h/common/gsl-1.15/include -I/h/common/BUFRLIB_v10-0-1
>>> -I/h/common/libf2c \
>>> -L/home/qcteam/METv4.1/lib
>>> -L/home/qcteam/netcdf/netcdf-4.1.3/compiled/lib
>>> -L/h/common/gsl-1.15/lib -L/h/common/BUFRLIB_v10-0-1
>>> -L/h/common/libf2c \
>>> -lvx_stat_out \
>>> -lvx_statistics \
>>> -lvx_shapedata \
>>> -lvx_gsl_prob \
>>> -lvx_analysis_util \
>>> -lvx_data2d_grib \
>>> -lvx_data2d \
>>> -lvx_nc_util \
>>> -lvx_grid \
>>> -lvx_config \
>>> -lvx_pb_util \
>>> -lvx_cal \
>>> -lvx_util \
>>> -lvx_math \
>>> -lvx_color \
>>> -lvx_log \
>>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas -lbufr \
>>> -lf2c -lgfortran
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:70: error: \u2018::isalnum\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:71: error: \u2018::isalpha\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:72: error: \u2018::iscntrl\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:73: error: \u2018::isdigit\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:74: error: \u2018::isgraph\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:75: error: \u2018::islower\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:76: error: \u2018::isprint\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:77: error: \u2018::ispunct\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:78: error: \u2018::isspace\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:79: error: \u2018::isupper\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:80: error: \u2018::isxdigit\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:81: error: \u2018::tolower\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> c
>>> ctype:82: error: \u2018::toupper\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:51: error:
>>> \u2018_ISupper\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:52: error:
>>> \u2018_ISlower\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:53: error:
>>> \u2018_ISalpha\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:54: error:
>>> \u2018_ISdigit\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:55: error:
>>> \u2018_ISxdigit\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:56: error:
>>> \u2018_ISspace\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:57: error:
>>> \u2018_ISprint\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
>>> \u2018_ISalpha\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
>>> \u2018_ISdigit\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
>>> \u2018_ISpunct\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:59: error:
>>> \u2018_IScntrl\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:60: error:
>>> \u2018_ISpunct\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
>>> \u2018_ISalpha\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2/
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
>>> \u2018_ISdigit\u2019
>>> was not declared in this scope
>>> make[4]: *** [pb2nc] Error 1
>>> make[4]: Leaving directory
>>> `/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>>> make[3]: *** [all] Error 2
>>> make[3]: Leaving directory
>>> `/ibrix/home/qcteam/METv4.1/src/tools/other'
>>> make[2]: *** [all] Error 2
>>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>> make[1]: *** [all] Error 2
>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>> make: *** [all] Error 2
>>>
>>> -----Original Message-----
>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>> Sent: Wednesday, June 12, 2013 12:08 PM
>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>
>>> Matt,
>>>
>>> The answer is in the error message you sent me:
>>> /usr/bin/ld: skipping incompatible
>>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>>> -lnetcdf_c++
>>>
>>> The linker isn't using the version of the NetCDF library you told
>>> it to because it is "incompatible". In my experience, that either
>>> means that that library was built using a different family of
>>> compilers (e.g. PGI vs GNU) or that it was built using an earlier
>>> version of the GNU compilers. I suspect that if you tried to
rebuild
>>> METv4.0 right now with the current version of the GNU compilers,
>>> you'd see the same linker error.
>>>
>>> I'd suggest rebuilding the NetCDF library in /h/common/netcdf-
4.1.3
>>> using the current version of the GNU compilers and then try again.
>>>
>>> Hope that helps.
>>>
>>> Thanks,
>>> John
>>>
>>> On 06/12/2013 11:00 AM, matthew.sittel.ctr at offutt.af.mil via RT
>>> wrote:
>>>>
>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>>
>>>> OK, let me try again using the 4.1.3 we know works (at least for
>>>> v4.0). Here's where it fails during the MET v4.1 build:
>>>>
>>>> make[2]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools'
>>>> make[3]: Entering directory
>>>> `/ibrix/home/qcteam/METv4.1/src/tools/core'
>>>> make[4]: Entering directory
>>>> `/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>>>
>>>> /usr/bin/g++ -m32 ensemble_stat_conf_info.cc -Wall -Wshadow
>>>> -static -g -c -I/home/qcteam/METv4.1/include
>>>> -I/h/common/gsl-1.15/include -I/h/common/netcdf-4.1.3/include
>>>> /usr/bin/g++ -m32 -o ensemble_stat ensemble_stat.cc
>>>> ensemble_stat_conf_info.o \
>>>> -Wall -Wshadow -static -g
>>>> -DMET_BASE=\"/home/qcteam/METv4.1\" \
>>>> -I/home/qcteam/METv4.1/include
>>>> -I/h/common/netcdf-4.1.3/include -I/h/common/gsl-1.15/include \
>>>> -L/home/qcteam/METv4.1/lib -L/h/common/netcdf-
4.1.3/lib
>>>> -L/h/common/gsl-1.15/lib \
>>>> -lvx_stat_out \
>>>> -lvx_statistics \
>>>> -lvx_shapedata \
>>>> -lvx_gsl_prob \
>>>> -lvx_analysis_util \
>>>> -lvx_data2d_factory \
>>>> -lvx_data2d_nc_met \
>>>> -lvx_data2d_grib \
>>>> -lvx_data2d_nc_pinterp \
>>>> -lvx_data2d_nccf \
>>>> -lvx_data2d \
>>>> -lvx_nc_util \
>>>> -lvx_grid \
>>>> -lvx_config \
>>>> -lvx_cal \
>>>> -lvx_util \
>>>> -lvx_math \
>>>> -lvx_color \
>>>> -lvx_log \
>>>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas \
>>>>
>>>> /usr/bin/ld: skipping incompatible
>>>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>>>> -lnetcdf_c++
>>>> /usr/bin/ld: cannot find -lnetcdf_c++
>>>> collect2: ld returned 1 exit status
>>>> make[4]: *** [ensemble_stat] Error 1
>>>> make[4]: Leaving directory
>>>> `/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>>> make[3]: *** [all] Error 2
>>>> make[3]: Leaving directory
>>>> `/ibrix/home/qcteam/METv4.1/src/tools/core'
>>>> make[2]: *** [all] Error 2
>>>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>>> make[1]: *** [all] Error 2
>>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>>> make: *** [all] Error 2
>>>>
>>>> -----Original Message-----
>>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>>> Sent: Wednesday, June 12, 2013 11:43 AM
>>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>>
>>>> Matt,
>>>>
>>>> We've typically been compiling MET using NetCDF version 3 - the
>>>> specific version is netcdf-3.6.3.
>>>>
>>>> However, I just downloaded netcdf-4.1.3 and compiled it with the
>>>> GNU compilers (gcc and g++) using the same options you did
>>>> (--disable-netcdf-4 --disable-dap). Then I recompiled METv4.1
using
>>>> netcdf-4.1.3, and it worked fine.
>>>>
>>>> Your netcdf build definitely needs to have include/netcdf.hh - so
>>>> that why 4.3.0 wasn't working.
>>>>
>>>> Can you please do a 'make clean' followed by 'make >&
>>>> make_met.log' when compiling with netcdf-4.1.3 and then send me
the
>>>> log file? What type of compilers are you using? Also, it's
>>>> possible that if the version of the compiler has changed since
the
>>>> time you compiled netcdf-4.1.3 to now, that could be causing a
>>>> problem.
>>>>
>>>> Thanks,
>>>> John
>>>>
>>>> On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT
>>>> wrote:
>>>>>
>>>>> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
>>>>> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
>>>>> Queue: met_help
>>>>> Subject: NETCDF Issue
>>>>> Owner: Nobody
>>>>> Requestors: matthew.sittel.ctr at offutt.af.mil
>>>>> Status: new
>>>>> Ticket <URL:
>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830
>>>>>>
>>>>>
>>>>>
>>>>> Hi,
>>>>>
>>>>> Okay, now I've *really* got a problem with the compile that I
>>>>> can't solve.
>>>>>
>>>>> Which version of NetCDF did you use to get METv4.1 to compile?
>>>>> The compile I am running here at AFWA failed for version 4.1.3
>>>>> (4.1.3 works great for the compile of METv4.0 though), and also
>>>>> fails for the latest version of netcdf, 4.3.0. With the latter
>>>>> case, here's where the compilation fails:
>>>>>
>>>>> *** Finished Making the vx_grid Library ***
>>>>>
>>>>> make[3]: Leaving directory
>>>>> `/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
>>>>> make[3]: Entering directory
>>>>> `/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
>>>>> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
>>>>> -I/home/qcteam/METv4.1/include
>>>>> -I/home/qcteam/netcdf/compiled/include
>>>>> In file included from nc_var_info.cc:24:
>>>>> nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
>>>>>
>>>>> The nectdf 'include' I compiled for 4.3.0 has only one file in
>>>>> it, netcdf.h, while 4.1.3 has a few more, including the
netcdf.hh
>>>>> that adds the inclusion of one more file from the 4.1.3 include
>>>>> (netcdfcpp.h).
>>>>>
>>>>> The routine say netcdf compiled successfully for me, but I had
to
>>>>> disable a couple of things to get it to run correctly: (I ran
the
>>>>> following: "./configure --prefix=/home/qcteam/netcdf/compiled
>>>>> --disable-netcdf-4 --disable-dap"). Is it possible one of those
>>>>> disables is the reason for the missing file(s)?
>>>>>
>>>>> It looks as if the timing of the 4.1 release might be using an
>>>>> older netcdf than 4.3.0? Rather than try every iteration of
netcdf
>>>>> between 4.1.3 and 4.3.0, I thought I'd ask. I wasn't going to
>>>>> update netcdf but the compile of 'pb2nc' while using 4.1.3
failed,
>>>>> suggesting I had to update.
>>>>>
>>>>> Thanks,
>>>>> Matt
>>>>>
>>>>> // SIGNED //
>>>>> Matthew C. Sittel
>>>>> Associate Scientist
>>>>> University Corporation for Atmospheric Research 16WS/WXN, Offutt
>>>>> AFB, NE
>>>>> (402) 294-3473 DSN: 271-3473
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #61830] NETCDF Issue
From: Craig, Robert J Civ USAF AFWA 16 WS/WXN
Time: Thu Jun 13 14:01:51 2013
John, the bug fix was one program. Where do I put it and do I need to
recompile MET?
Thanks
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Thursday, June 13, 2013 2:34 PM
To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
Subject: RE: [rt.rap.ucar.edu #61830] NETCDF Issue
Bob,
I've posted a bugfix for METv4.1 here:
http://www.dtcenter.org/met/users/support/known_issues/METv4.1/index.php
The change should allow you to use that GRIB file you sent me without
having to change it's suffix or use the file_type option in the config
file.
Please try applying the patch and let me know that they work for you.
Thanks,
John
On 2013-06-12 14:53, Craig, Robert J Civ USAF AFWA 16 WS/WXN via RT
wrote:
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>
> John, now that Matt has the software compiled, I tried to run my
mode
> program. I get the following error:
>
> /home/qcteam/METv4.1/bin/mode
> /home/qcteam/fcsts_t001_2013061100_000.GR1
> /h/data/global/WXQC/data/temp/met/cloud_obs_file_71
> /home/qcteam/METv4.1/data/config/DCFModeConfig_55 -outdir
> /home/qcteam/METv4.1/out/mode/ -v 10 DEBUG 1: Default Config File:
> /home/qcteam/METv4.1/data/config/MODEConfig_default
> DEBUG 1: Match Config File:
> /home/qcteam/METv4.1/data/config/DCFModeConfig_55
> DEBUG 1: Merge Config File:
> /home/qcteam/METv4.1/data/config/DCFModeConfig_55
> DEBUG 4: Met2dDataFileFactory::new_met_2d_data_file() -> created new
> Met2dDataFile object of type "FileType_None".
> ERROR :
> ERROR : Trouble reading forecast file
> "/home/qcteam/fcsts_t001_2013061100_000.GR1"
> ERROR :
>
> A degrib of the file header is attached (gribrecords). The file is
> wide open on read permissions. I ftp the file to your site -
> craig_data. I did try TCDC instead of 71 in the grib name field but
> got the same error.
>
> Thanks
>
> Bob
>
>
>
>
>
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Wednesday, June 12, 2013 3:01 PM
> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>
> Matt,
>
> If you don't have the HDF and HDF-EOS libraries set up, the MODIS
> piece won't compile.
>
> Regarding the problem with compiling tc_utils, I think you're
missing
> the "-fno-second-underscore" option in the FC_FLAGS line. It's
> included in the default user_defs_gnu.mk, but for some reason it
> appears not to be present in yours. Please try adding that back in,
> enable the compilation of the tc_utils, and try again.
>
> Thanks,
> John
>
> On 06/12/2013 01:26 PM, matthew.sittel.ctr at offutt.af.mil via RT
> wrote:
>>
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>
>> I got through that step after removing the F2C settings, as you
>> suggested.
>>
>> I got everything else to compile *if* I left out modis_regrid and
the
>> tropical steps (i.e., set those last two options to 0 in the mk
>> file).
>>
>> Modis mentioned something about "hdf", which I think is part of the
>> NetCDF routine-I recall reading that in the "INSTALL" part of
NetCDF.
>>
>> The error I got from the tropical part might be useful to you?
>>
>> ===
>> make[4]: Entering directory
>> `/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
>>
>> /usr/bin/gfortran acerr.f -Wall -Wshadow -static -ff2c -g -c
>>
>> /usr/bin/gfortran clip_shif.f -Wall -Wshadow -static -ff2c -g -c >
>> /dev/null 2>&1
>>
>> /usr/bin/g++ tc_pairs_conf_info.cc -Wall -Wshadow -static -g -c
>> -I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
>> -I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>>
>> /usr/bin/g++ -o tc_pairs tc_pairs.cc acerr.o clip_shif.o
>> tc_pairs_conf_info.o \
>> -Wall -Wshadow -static -g
>> -DMET_BASE=\"/home/qcteam/METv4.1\" \
>> -I/home/qcteam/METv4.1/include
>> -I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>> -I/h/common/gsl-1.15/include \
>> -L/home/qcteam/METv4.1/lib
>> -L/home/qcteam/netcdf/netcdf-4.1.3/compiled/lib
>> -L/h/common/gsl-1.15/lib \
>> -lvx_config \
>> -lvx_data2d_nc_met \
>> -lvx_analysis_util \
>> -lvx_tc_util \
>> -lvx_nc_util \
>> -lvx_grid \
>> -lvx_config \
>> -lvx_util \
>> -lvx_math \
>> -lvx_cal \
>> -lvx_log \
>> -lm \
>> -lnetcdf_c++ -lnetcdf \
>> -lgfortran
>> /tmp/ccJGmQPZ.o: In function `derive_baseline_model':
>> /home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs/tc_pairs.cc:1229:
>> undefined reference to `oclipd5_test_'
>> collect2: ld returned 1 exit status
>> make[4]: *** [tc_pairs] Error 1
>> make[4]: Leaving directory
>> `/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
>> make[3]: *** [all] Error 2
>> make[3]: Leaving directory
>> `/ibrix/home/qcteam/METv4.1/src/tools/tc_utils'
>> make[2]: *** [all] Error 2
>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>> make[1]: *** [all] Error 2
>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>> make: *** [all] Error 2
>>
>> -----Original Message-----
>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>> Sent: Wednesday, June 12, 2013 1:58 PM
>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>
>> Matt,
>>
>> I read through some previous MET-Help questions and have a
suggestion
>> for you to try - in user_defs.mk just empty out all of the
"F2C_..."
>> settings:
>> F2C_BASE =
>> F2C_INCS =
>> F2C_LIBS =
>> F2C_LIBNAME =
>>
>> Based on the log output you sent, it looks like you've specified
>> something in those settings. That's only necessary for the old GNU
>> g77, and you're using gfortran.
>>
>> Please let me know if that does the trick.
>>
>> Thanks,
>> John
>>
>> On 06/12/2013 12:11 PM, matthew.sittel.ctr at offutt.af.mil via RT
>> wrote:
>>>
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>
>>> Hey John,
>>>
>>> Okay, I re-compiled netcdf 4.1.3 successfully, and sourced it.
>>> Here’s the error I get now. Since I get the same errors when
>>> compiling with the older netcdf we already had installed and
>>> compiled in 2011, I think we can rule out netcdf as the issue.
>>>
>>> =====
>>>
>>> make[4]: Entering directory
>>> `/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>>>
>>> /usr/bin/g++ pb2nc_conf_info.cc -Wall -Wshadow -static -g -c
>>> -I/home/qcteam/METv4.1/include
>>>
>>> /usr/bin/gfortran numpbmsg.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran openpb.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran closepb.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran readpb.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran dumppb.f -Wall -Wshadow -static -ff2c -g -c
>>> /usr/bin/g++ -o pb2nc pb2nc.cc pb2nc_conf_info.o numpbmsg.o
>>> openpb.o closepb.o readpb.o dumppb.o \
>>> -Wall -Wshadow -static -g
>>> -DMET_BASE=\"/home/qcteam/METv4.1\" \
>>> -I/home/qcteam/METv4.1/include
>>> -I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>>> -I/h/common/gsl-1.15/include -I/h/common/BUFRLIB_v10-0-1
>>> -I/h/common/libf2c \
>>> -L/home/qcteam/METv4.1/lib
>>> -L/home/qcteam/netcdf/netcdf-4.1.3/compiled/lib
>>> -L/h/common/gsl-1.15/lib -L/h/common/BUFRLIB_v10-0-1
>>> -L/h/common/libf2c \
>>> -lvx_stat_out \
>>> -lvx_statistics \
>>> -lvx_shapedata \
>>> -lvx_gsl_prob \
>>> -lvx_analysis_util \
>>> -lvx_data2d_grib \
>>> -lvx_data2d \
>>> -lvx_nc_util \
>>> -lvx_grid \
>>> -lvx_config \
>>> -lvx_pb_util \
>>> -lvx_cal \
>>> -lvx_util \
>>> -lvx_math \
>>> -lvx_color \
>>> -lvx_log \
>>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas -lbufr \
>>> -lf2c -lgfortran
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> c
>>> ctype:70: error: \u2018::isalnum\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> c
>>> ctype:71: error: \u2018::isalpha\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> c
>>> ctype:72: error: \u2018::iscntrl\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> c
>>> ctype:73: error: \u2018::isdigit\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> c
>>> ctype:74: error: \u2018::isgraph\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> c
>>> ctype:75: error: \u2018::islower\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> c
>>> ctype:76: error: \u2018::isprint\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> c
>>> ctype:77: error: \u2018::ispunct\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> c
>>> ctype:78: error: \u2018::isspace\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> c
>>> ctype:79: error: \u2018::isupper\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> c
>>> ctype:80: error: \u2018::isxdigit\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> c
>>> ctype:81: error: \u2018::tolower\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> c
>>> ctype:82: error: \u2018::toupper\u2019 has not been declared
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:51: error:
>>> \u2018_ISupper\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:52: error:
>>> \u2018_ISlower\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:53: error:
>>> \u2018_ISalpha\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:54: error:
>>> \u2018_ISdigit\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:55: error:
>>> \u2018_ISxdigit\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:56: error:
>>> \u2018_ISspace\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:57: error:
>>> \u2018_ISprint\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
>>> \u2018_ISalpha\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
>>> \u2018_ISdigit\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
>>> \u2018_ISpunct\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:59: error:
>>> \u2018_IScntrl\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:60: error:
>>> \u2018_ISpunct\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
>>> \u2018_ISalpha\u2019
>>> was not declared in this scope
>>>
>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>> /
>>> x
>>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
>>> \u2018_ISdigit\u2019
>>> was not declared in this scope
>>> make[4]: *** [pb2nc] Error 1
>>> make[4]: Leaving directory
>>> `/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>>> make[3]: *** [all] Error 2
>>> make[3]: Leaving directory
>>> `/ibrix/home/qcteam/METv4.1/src/tools/other'
>>> make[2]: *** [all] Error 2
>>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>> make[1]: *** [all] Error 2
>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>> make: *** [all] Error 2
>>>
>>> -----Original Message-----
>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>> Sent: Wednesday, June 12, 2013 12:08 PM
>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>
>>> Matt,
>>>
>>> The answer is in the error message you sent me:
>>> /usr/bin/ld: skipping incompatible
>>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>>> -lnetcdf_c++
>>>
>>> The linker isn't using the version of the NetCDF library you told
it
>>> to because it is "incompatible". In my experience, that either
>>> means that that library was built using a different family of
>>> compilers (e.g. PGI vs GNU) or that it was built using an earlier
>>> version of the GNU compilers. I suspect that if you tried to
>>> rebuild
>>> METv4.0 right now with the current version of the GNU compilers,
>>> you'd see the same linker error.
>>>
>>> I'd suggest rebuilding the NetCDF library in /h/common/netcdf-
4.1.3
>>> using the current version of the GNU compilers and then try again.
>>>
>>> Hope that helps.
>>>
>>> Thanks,
>>> John
>>>
>>> On 06/12/2013 11:00 AM, matthew.sittel.ctr at offutt.af.mil via RT
>>> wrote:
>>>>
>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>>
>>>> OK, let me try again using the 4.1.3 we know works (at least for
>>>> v4.0). Here's where it fails during the MET v4.1 build:
>>>>
>>>> make[2]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools'
>>>> make[3]: Entering directory
>>>> `/ibrix/home/qcteam/METv4.1/src/tools/core'
>>>> make[4]: Entering directory
>>>> `/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>>>
>>>> /usr/bin/g++ -m32 ensemble_stat_conf_info.cc -Wall -Wshadow
>>>> -static -g -c -I/home/qcteam/METv4.1/include
>>>> -I/h/common/gsl-1.15/include -I/h/common/netcdf-4.1.3/include
>>>> /usr/bin/g++ -m32 -o ensemble_stat ensemble_stat.cc
>>>> ensemble_stat_conf_info.o \
>>>> -Wall -Wshadow -static -g
>>>> -DMET_BASE=\"/home/qcteam/METv4.1\" \
>>>> -I/home/qcteam/METv4.1/include
>>>> -I/h/common/netcdf-4.1.3/include -I/h/common/gsl-1.15/include \
>>>> -L/home/qcteam/METv4.1/lib -L/h/common/netcdf-
4.1.3/lib
>>>> -L/h/common/gsl-1.15/lib \
>>>> -lvx_stat_out \
>>>> -lvx_statistics \
>>>> -lvx_shapedata \
>>>> -lvx_gsl_prob \
>>>> -lvx_analysis_util \
>>>> -lvx_data2d_factory \
>>>> -lvx_data2d_nc_met \
>>>> -lvx_data2d_grib \
>>>> -lvx_data2d_nc_pinterp \
>>>> -lvx_data2d_nccf \
>>>> -lvx_data2d \
>>>> -lvx_nc_util \
>>>> -lvx_grid \
>>>> -lvx_config \
>>>> -lvx_cal \
>>>> -lvx_util \
>>>> -lvx_math \
>>>> -lvx_color \
>>>> -lvx_log \
>>>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas \
>>>>
>>>> /usr/bin/ld: skipping incompatible
>>>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>>>> -lnetcdf_c++
>>>> /usr/bin/ld: cannot find -lnetcdf_c++
>>>> collect2: ld returned 1 exit status
>>>> make[4]: *** [ensemble_stat] Error 1
>>>> make[4]: Leaving directory
>>>> `/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>>> make[3]: *** [all] Error 2
>>>> make[3]: Leaving directory
>>>> `/ibrix/home/qcteam/METv4.1/src/tools/core'
>>>> make[2]: *** [all] Error 2
>>>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>>> make[1]: *** [all] Error 2
>>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>>> make: *** [all] Error 2
>>>>
>>>> -----Original Message-----
>>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>>> Sent: Wednesday, June 12, 2013 11:43 AM
>>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>>
>>>> Matt,
>>>>
>>>> We've typically been compiling MET using NetCDF version 3 - the
>>>> specific version is netcdf-3.6.3.
>>>>
>>>> However, I just downloaded netcdf-4.1.3 and compiled it with the
>>>> GNU compilers (gcc and g++) using the same options you did
>>>> (--disable-netcdf-4 --disable-dap). Then I recompiled METv4.1
>>>> using netcdf-4.1.3, and it worked fine.
>>>>
>>>> Your netcdf build definitely needs to have include/netcdf.hh - so
>>>> that why 4.3.0 wasn't working.
>>>>
>>>> Can you please do a 'make clean' followed by 'make >&
make_met.log'
>>>> when compiling with netcdf-4.1.3 and then send me the log file?
>>>> What type of compilers are you using? Also, it's possible that
if
>>>> the version of the compiler has changed since the time you
compiled
>>>> netcdf-4.1.3 to now, that could be causing a problem.
>>>>
>>>> Thanks,
>>>> John
>>>>
>>>> On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT
>>>> wrote:
>>>>>
>>>>> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
>>>>> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
>>>>> Queue: met_help
>>>>> Subject: NETCDF Issue
>>>>> Owner: Nobody
>>>>> Requestors: matthew.sittel.ctr at offutt.af.mil
>>>>> Status: new
>>>>> Ticket <URL:
>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830
>>>>>>
>>>>>
>>>>>
>>>>> Hi,
>>>>>
>>>>> Okay, now I've *really* got a problem with the compile that I
>>>>> can't solve.
>>>>>
>>>>> Which version of NetCDF did you use to get METv4.1 to compile?
>>>>> The compile I am running here at AFWA failed for version 4.1.3
>>>>> (4.1.3 works great for the compile of METv4.0 though), and also
>>>>> fails for the latest version of netcdf, 4.3.0. With the latter
>>>>> case, here's where the compilation fails:
>>>>>
>>>>> *** Finished Making the vx_grid Library ***
>>>>>
>>>>> make[3]: Leaving directory
>>>>> `/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
>>>>> make[3]: Entering directory
>>>>> `/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
>>>>> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
>>>>> -I/home/qcteam/METv4.1/include
>>>>> -I/home/qcteam/netcdf/compiled/include
>>>>> In file included from nc_var_info.cc:24:
>>>>> nc_var_info.h:24:21: error: netcdf.hh: No such file or directory
>>>>>
>>>>> The nectdf 'include' I compiled for 4.3.0 has only one file in
it,
>>>>> netcdf.h, while 4.1.3 has a few more, including the netcdf.hh
that
>>>>> adds the inclusion of one more file from the 4.1.3 include
>>>>> (netcdfcpp.h).
>>>>>
>>>>> The routine say netcdf compiled successfully for me, but I had
to
>>>>> disable a couple of things to get it to run correctly: (I ran
the
>>>>> following: "./configure --prefix=/home/qcteam/netcdf/compiled
>>>>> --disable-netcdf-4 --disable-dap"). Is it possible one of those
>>>>> disables is the reason for the missing file(s)?
>>>>>
>>>>> It looks as if the timing of the 4.1 release might be using an
>>>>> older netcdf than 4.3.0? Rather than try every iteration of
>>>>> netcdf between 4.1.3 and 4.3.0, I thought I'd ask. I wasn't
going
>>>>> to update netcdf but the compile of 'pb2nc' while using 4.1.3
>>>>> failed, suggesting I had to update.
>>>>>
>>>>> Thanks,
>>>>> Matt
>>>>>
>>>>> // SIGNED //
>>>>> Matthew C. Sittel
>>>>> Associate Scientist
>>>>> University Corporation for Atmospheric Research 16WS/WXN, Offutt
>>>>> AFB, NE
>>>>> (402) 294-3473 DSN: 271-3473
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #61830] NETCDF Issue
From: John Halley Gotway
Time: Thu Jun 13 14:27:06 2013
Bob,
Just follow the instructions listed at the top of that page:
http://www.dtcenter.org/met/users/support/known_issues/METv4.1/index.php
Retrieve all of the recommended updates for METv4.1 in a tarfile:
- Download the tarfile METv4.1_patches_20130613.tar.gz
- Copy it into the top-level METv4.1 directory.
- Uncompress and untar the file. Please note that the new version
of
the files in the tarball will overwrite the original version of those
files. Any changes you may have made will be lost.
- Rebuild MET, being sure perform a make clean first.
This will update the library code that reads the gridded data files.
Just let me know if you get stuck.
Thanks,
John
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>
> John, the bug fix was one program. Where do I put it and do I need
to
> recompile MET?
>
> Thanks
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Thursday, June 13, 2013 2:34 PM
> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Subject: RE: [rt.rap.ucar.edu #61830] NETCDF Issue
>
> Bob,
>
> I've posted a bugfix for METv4.1 here:
>
>
http://www.dtcenter.org/met/users/support/known_issues/METv4.1/index.php
>
> The change should allow you to use that GRIB file you sent me
without
> having to change it's suffix or use the file_type option in the
config
> file.
>
> Please try applying the patch and let me know that they work for
you.
>
> Thanks,
> John
>
>
> On 2013-06-12 14:53, Craig, Robert J Civ USAF AFWA 16 WS/WXN via RT
> wrote:
>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>
>> John, now that Matt has the software compiled, I tried to run my
mode
>> program. I get the following error:
>>
>> /home/qcteam/METv4.1/bin/mode
>> /home/qcteam/fcsts_t001_2013061100_000.GR1
>> /h/data/global/WXQC/data/temp/met/cloud_obs_file_71
>> /home/qcteam/METv4.1/data/config/DCFModeConfig_55 -outdir
>> /home/qcteam/METv4.1/out/mode/ -v 10 DEBUG 1: Default Config File:
>> /home/qcteam/METv4.1/data/config/MODEConfig_default
>> DEBUG 1: Match Config File:
>> /home/qcteam/METv4.1/data/config/DCFModeConfig_55
>> DEBUG 1: Merge Config File:
>> /home/qcteam/METv4.1/data/config/DCFModeConfig_55
>> DEBUG 4: Met2dDataFileFactory::new_met_2d_data_file() -> created
new
>> Met2dDataFile object of type "FileType_None".
>> ERROR :
>> ERROR : Trouble reading forecast file
>> "/home/qcteam/fcsts_t001_2013061100_000.GR1"
>> ERROR :
>>
>> A degrib of the file header is attached (gribrecords). The file is
>> wide open on read permissions. I ftp the file to your site -
>> craig_data. I did try TCDC instead of 71 in the grib name field
but
>> got the same error.
>>
>> Thanks
>>
>> Bob
>>
>>
>>
>>
>>
>>
>> -----Original Message-----
>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>> Sent: Wednesday, June 12, 2013 3:01 PM
>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>
>> Matt,
>>
>> If you don't have the HDF and HDF-EOS libraries set up, the MODIS
>> piece won't compile.
>>
>> Regarding the problem with compiling tc_utils, I think you're
missing
>> the "-fno-second-underscore" option in the FC_FLAGS line. It's
>> included in the default user_defs_gnu.mk, but for some reason it
>> appears not to be present in yours. Please try adding that back
in,
>> enable the compilation of the tc_utils, and try again.
>>
>> Thanks,
>> John
>>
>> On 06/12/2013 01:26 PM, matthew.sittel.ctr at offutt.af.mil via RT
>> wrote:
>>>
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>
>>> I got through that step after removing the F2C settings, as you
>>> suggested.
>>>
>>> I got everything else to compile *if* I left out modis_regrid and
the
>>> tropical steps (i.e., set those last two options to 0 in the mk
>>> file).
>>>
>>> Modis mentioned something about "hdf", which I think is part of
the
>>> NetCDF routine-I recall reading that in the "INSTALL" part of
NetCDF.
>>>
>>> The error I got from the tropical part might be useful to you?
>>>
>>> ===
>>> make[4]: Entering directory
>>> `/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
>>>
>>> /usr/bin/gfortran acerr.f -Wall -Wshadow -static -ff2c -g -c
>>>
>>> /usr/bin/gfortran clip_shif.f -Wall -Wshadow -static -ff2c -g -c
>
>>> /dev/null 2>&1
>>>
>>> /usr/bin/g++ tc_pairs_conf_info.cc -Wall -Wshadow -static -g
-c
>>> -I/home/qcteam/METv4.1/include -I/h/common/gsl-1.15/include
>>> -I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>>>
>>> /usr/bin/g++ -o tc_pairs tc_pairs.cc acerr.o clip_shif.o
>>> tc_pairs_conf_info.o \
>>> -Wall -Wshadow -static -g
>>> -DMET_BASE=\"/home/qcteam/METv4.1\" \
>>> -I/home/qcteam/METv4.1/include
>>> -I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>>> -I/h/common/gsl-1.15/include \
>>> -L/home/qcteam/METv4.1/lib
>>> -L/home/qcteam/netcdf/netcdf-4.1.3/compiled/lib
>>> -L/h/common/gsl-1.15/lib \
>>> -lvx_config \
>>> -lvx_data2d_nc_met \
>>> -lvx_analysis_util \
>>> -lvx_tc_util \
>>> -lvx_nc_util \
>>> -lvx_grid \
>>> -lvx_config \
>>> -lvx_util \
>>> -lvx_math \
>>> -lvx_cal \
>>> -lvx_log \
>>> -lm \
>>> -lnetcdf_c++ -lnetcdf \
>>> -lgfortran
>>> /tmp/ccJGmQPZ.o: In function `derive_baseline_model':
>>> /home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs/tc_pairs.cc:1229:
>>> undefined reference to `oclipd5_test_'
>>> collect2: ld returned 1 exit status
>>> make[4]: *** [tc_pairs] Error 1
>>> make[4]: Leaving directory
>>> `/ibrix/home/qcteam/METv4.1/src/tools/tc_utils/tc_pairs'
>>> make[3]: *** [all] Error 2
>>> make[3]: Leaving directory
>>> `/ibrix/home/qcteam/METv4.1/src/tools/tc_utils'
>>> make[2]: *** [all] Error 2
>>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>> make[1]: *** [all] Error 2
>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>> make: *** [all] Error 2
>>>
>>> -----Original Message-----
>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>> Sent: Wednesday, June 12, 2013 1:58 PM
>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>
>>> Matt,
>>>
>>> I read through some previous MET-Help questions and have a
suggestion
>>> for you to try - in user_defs.mk just empty out all of the
"F2C_..."
>>> settings:
>>> F2C_BASE =
>>> F2C_INCS =
>>> F2C_LIBS =
>>> F2C_LIBNAME =
>>>
>>> Based on the log output you sent, it looks like you've specified
>>> something in those settings. That's only necessary for the old
GNU
>>> g77, and you're using gfortran.
>>>
>>> Please let me know if that does the trick.
>>>
>>> Thanks,
>>> John
>>>
>>> On 06/12/2013 12:11 PM, matthew.sittel.ctr at offutt.af.mil via RT
>>> wrote:
>>>>
>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>>
>>>> Hey John,
>>>>
>>>> Okay, I re-compiled netcdf 4.1.3 successfully, and sourced it.
>>>> Here’s the error I get now. Since I get the same errors when
>>>> compiling with the older netcdf we already had installed and
>>>> compiled in 2011, I think we can rule out netcdf as the issue.
>>>>
>>>> =====
>>>>
>>>> make[4]: Entering directory
>>>> `/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>>>>
>>>> /usr/bin/g++ pb2nc_conf_info.cc -Wall -Wshadow -static -g -c
>>>> -I/home/qcteam/METv4.1/include
>>>>
>>>> /usr/bin/gfortran numpbmsg.f -Wall -Wshadow -static -ff2c -g -c
>>>>
>>>> /usr/bin/gfortran openpb.f -Wall -Wshadow -static -ff2c -g -c
>>>>
>>>> /usr/bin/gfortran closepb.f -Wall -Wshadow -static -ff2c -g -c
>>>>
>>>> /usr/bin/gfortran readpb.f -Wall -Wshadow -static -ff2c -g -c
>>>>
>>>> /usr/bin/gfortran dumppb.f -Wall -Wshadow -static -ff2c -g -c
>>>> /usr/bin/g++ -o pb2nc pb2nc.cc pb2nc_conf_info.o numpbmsg.o
>>>> openpb.o closepb.o readpb.o dumppb.o \
>>>> -Wall -Wshadow -static -g
>>>> -DMET_BASE=\"/home/qcteam/METv4.1\" \
>>>> -I/home/qcteam/METv4.1/include
>>>> -I/home/qcteam/netcdf/netcdf-4.1.3/compiled/include
>>>> -I/h/common/gsl-1.15/include -I/h/common/BUFRLIB_v10-0-1
>>>> -I/h/common/libf2c \
>>>> -L/home/qcteam/METv4.1/lib
>>>> -L/home/qcteam/netcdf/netcdf-4.1.3/compiled/lib
>>>> -L/h/common/gsl-1.15/lib -L/h/common/BUFRLIB_v10-0-1
>>>> -L/h/common/libf2c \
>>>> -lvx_stat_out \
>>>> -lvx_statistics \
>>>> -lvx_shapedata \
>>>> -lvx_gsl_prob \
>>>> -lvx_analysis_util \
>>>> -lvx_data2d_grib \
>>>> -lvx_data2d \
>>>> -lvx_nc_util \
>>>> -lvx_grid \
>>>> -lvx_config \
>>>> -lvx_pb_util \
>>>> -lvx_cal \
>>>> -lvx_util \
>>>> -lvx_math \
>>>> -lvx_color \
>>>> -lvx_log \
>>>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas -lbufr \
>>>> -lf2c -lgfortran
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> c
>>>> ctype:70: error: \u2018::isalnum\u2019 has not been declared
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> c
>>>> ctype:71: error: \u2018::isalpha\u2019 has not been declared
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> c
>>>> ctype:72: error: \u2018::iscntrl\u2019 has not been declared
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> c
>>>> ctype:73: error: \u2018::isdigit\u2019 has not been declared
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> c
>>>> ctype:74: error: \u2018::isgraph\u2019 has not been declared
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> c
>>>> ctype:75: error: \u2018::islower\u2019 has not been declared
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> c
>>>> ctype:76: error: \u2018::isprint\u2019 has not been declared
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> c
>>>> ctype:77: error: \u2018::ispunct\u2019 has not been declared
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> c
>>>> ctype:78: error: \u2018::isspace\u2019 has not been declared
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> c
>>>> ctype:79: error: \u2018::isupper\u2019 has not been declared
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> c
>>>> ctype:80: error: \u2018::isxdigit\u2019 has not been declared
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> c
>>>> ctype:81: error: \u2018::tolower\u2019 has not been declared
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> c
>>>> ctype:82: error: \u2018::toupper\u2019 has not been declared
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:51: error:
>>>> \u2018_ISupper\u2019
>>>> was not declared in this scope
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:52: error:
>>>> \u2018_ISlower\u2019
>>>> was not declared in this scope
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:53: error:
>>>> \u2018_ISalpha\u2019
>>>> was not declared in this scope
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:54: error:
>>>> \u2018_ISdigit\u2019
>>>> was not declared in this scope
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:55: error:
>>>> \u2018_ISxdigit\u2019
>>>> was not declared in this scope
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:56: error:
>>>> \u2018_ISspace\u2019
>>>> was not declared in this scope
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:57: error:
>>>> \u2018_ISprint\u2019
>>>> was not declared in this scope
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
>>>> \u2018_ISalpha\u2019
>>>> was not declared in this scope
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
>>>> \u2018_ISdigit\u2019
>>>> was not declared in this scope
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:58: error:
>>>> \u2018_ISpunct\u2019
>>>> was not declared in this scope
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:59: error:
>>>> \u2018_IScntrl\u2019
>>>> was not declared in this scope
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:60: error:
>>>> \u2018_ISpunct\u2019
>>>> was not declared in this scope
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
>>>> \u2018_ISalpha\u2019
>>>> was not declared in this scope
>>>>
>>>> /usr/lib/gcc/x86_64-redhat-
linux/4.1.2/../../../../include/c++/4.1.2
>>>> /
>>>> x
>>>> 86_64-redhat-linux/bits/ctype_base.h:61: error:
>>>> \u2018_ISdigit\u2019
>>>> was not declared in this scope
>>>> make[4]: *** [pb2nc] Error 1
>>>> make[4]: Leaving directory
>>>> `/ibrix/home/qcteam/METv4.1/src/tools/other/pb2nc'
>>>> make[3]: *** [all] Error 2
>>>> make[3]: Leaving directory
>>>> `/ibrix/home/qcteam/METv4.1/src/tools/other'
>>>> make[2]: *** [all] Error 2
>>>> make[2]: Leaving directory `/ibrix/home/qcteam/METv4.1/src/tools'
>>>> make[1]: *** [all] Error 2
>>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>>> make: *** [all] Error 2
>>>>
>>>> -----Original Message-----
>>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>>> Sent: Wednesday, June 12, 2013 12:08 PM
>>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>>
>>>> Matt,
>>>>
>>>> The answer is in the error message you sent me:
>>>> /usr/bin/ld: skipping incompatible
>>>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>>>> -lnetcdf_c++
>>>>
>>>> The linker isn't using the version of the NetCDF library you told
it
>>>> to because it is "incompatible". In my experience, that either
>>>> means that that library was built using a different family of
>>>> compilers (e.g. PGI vs GNU) or that it was built using an earlier
>>>> version of the GNU compilers. I suspect that if you tried to
>>>> rebuild
>>>> METv4.0 right now with the current version of the GNU compilers,
>>>> you'd see the same linker error.
>>>>
>>>> I'd suggest rebuilding the NetCDF library in /h/common/netcdf-
4.1.3
>>>> using the current version of the GNU compilers and then try
again.
>>>>
>>>> Hope that helps.
>>>>
>>>> Thanks,
>>>> John
>>>>
>>>> On 06/12/2013 11:00 AM, matthew.sittel.ctr at offutt.af.mil via RT
>>>> wrote:
>>>>>
>>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830 >
>>>>>
>>>>> OK, let me try again using the 4.1.3 we know works (at least for
>>>>> v4.0). Here's where it fails during the MET v4.1 build:
>>>>>
>>>>> make[2]: Entering directory
`/ibrix/home/qcteam/METv4.1/src/tools'
>>>>> make[3]: Entering directory
>>>>> `/ibrix/home/qcteam/METv4.1/src/tools/core'
>>>>> make[4]: Entering directory
>>>>> `/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>>>>
>>>>> /usr/bin/g++ -m32 ensemble_stat_conf_info.cc -Wall -Wshadow
>>>>> -static -g -c -I/home/qcteam/METv4.1/include
>>>>> -I/h/common/gsl-1.15/include -I/h/common/netcdf-4.1.3/include
>>>>> /usr/bin/g++ -m32 -o ensemble_stat ensemble_stat.cc
>>>>> ensemble_stat_conf_info.o \
>>>>> -Wall -Wshadow -static -g
>>>>> -DMET_BASE=\"/home/qcteam/METv4.1\" \
>>>>> -I/home/qcteam/METv4.1/include
>>>>> -I/h/common/netcdf-4.1.3/include -I/h/common/gsl-1.15/include \
>>>>> -L/home/qcteam/METv4.1/lib -L/h/common/netcdf-
4.1.3/lib
>>>>> -L/h/common/gsl-1.15/lib \
>>>>> -lvx_stat_out \
>>>>> -lvx_statistics \
>>>>> -lvx_shapedata \
>>>>> -lvx_gsl_prob \
>>>>> -lvx_analysis_util \
>>>>> -lvx_data2d_factory \
>>>>> -lvx_data2d_nc_met \
>>>>> -lvx_data2d_grib \
>>>>> -lvx_data2d_nc_pinterp \
>>>>> -lvx_data2d_nccf \
>>>>> -lvx_data2d \
>>>>> -lvx_nc_util \
>>>>> -lvx_grid \
>>>>> -lvx_config \
>>>>> -lvx_cal \
>>>>> -lvx_util \
>>>>> -lvx_math \
>>>>> -lvx_color \
>>>>> -lvx_log \
>>>>> -lm -lnetcdf_c++ -lnetcdf -lgsl -lgslcblas \
>>>>>
>>>>> /usr/bin/ld: skipping incompatible
>>>>> /h/common/netcdf-4.1.3/lib/libnetcdf_c++.a when searching for
>>>>> -lnetcdf_c++
>>>>> /usr/bin/ld: cannot find -lnetcdf_c++
>>>>> collect2: ld returned 1 exit status
>>>>> make[4]: *** [ensemble_stat] Error 1
>>>>> make[4]: Leaving directory
>>>>> `/ibrix/home/qcteam/METv4.1/src/tools/core/ensemble_stat'
>>>>> make[3]: *** [all] Error 2
>>>>> make[3]: Leaving directory
>>>>> `/ibrix/home/qcteam/METv4.1/src/tools/core'
>>>>> make[2]: *** [all] Error 2
>>>>> make[2]: Leaving directory
`/ibrix/home/qcteam/METv4.1/src/tools'
>>>>> make[1]: *** [all] Error 2
>>>>> make[1]: Leaving directory `/ibrix/home/qcteam/METv4.1/src'
>>>>> make: *** [all] Error 2
>>>>>
>>>>> -----Original Message-----
>>>>> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
>>>>> Sent: Wednesday, June 12, 2013 11:43 AM
>>>>> To: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
>>>>> Cc: Craig, Robert J Civ USAF AFWA 16 WS/WXN
>>>>> Subject: Re: [rt.rap.ucar.edu #61830] NETCDF Issue
>>>>>
>>>>> Matt,
>>>>>
>>>>> We've typically been compiling MET using NetCDF version 3 - the
>>>>> specific version is netcdf-3.6.3.
>>>>>
>>>>> However, I just downloaded netcdf-4.1.3 and compiled it with the
>>>>> GNU compilers (gcc and g++) using the same options you did
>>>>> (--disable-netcdf-4 --disable-dap). Then I recompiled METv4.1
>>>>> using netcdf-4.1.3, and it worked fine.
>>>>>
>>>>> Your netcdf build definitely needs to have include/netcdf.hh -
so
>>>>> that why 4.3.0 wasn't working.
>>>>>
>>>>> Can you please do a 'make clean' followed by 'make >&
make_met.log'
>>>>> when compiling with netcdf-4.1.3 and then send me the log file?
>>>>> What type of compilers are you using? Also, it's possible that
if
>>>>> the version of the compiler has changed since the time you
compiled
>>>>> netcdf-4.1.3 to now, that could be causing a problem.
>>>>>
>>>>> Thanks,
>>>>> John
>>>>>
>>>>> On 06/12/2013 10:05 AM, matthew.sittel.ctr at offutt.af.mil via RT
>>>>> wrote:
>>>>>>
>>>>>> Wed Jun 12 10:05:35 2013: Request 61830 was acted upon.
>>>>>> Transaction: Ticket created by matthew.sittel.ctr at offutt.af.mil
>>>>>> Queue: met_help
>>>>>> Subject: NETCDF Issue
>>>>>> Owner: Nobody
>>>>>> Requestors: matthew.sittel.ctr at offutt.af.mil
>>>>>> Status: new
>>>>>> Ticket <URL:
>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=61830
>>>>>>>
>>>>>>
>>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> Okay, now I've *really* got a problem with the compile that I
>>>>>> can't solve.
>>>>>>
>>>>>> Which version of NetCDF did you use to get METv4.1 to compile?
>>>>>> The compile I am running here at AFWA failed for version 4.1.3
>>>>>> (4.1.3 works great for the compile of METv4.0 though), and also
>>>>>> fails for the latest version of netcdf, 4.3.0. With the latter
>>>>>> case, here's where the compilation fails:
>>>>>>
>>>>>> *** Finished Making the vx_grid Library ***
>>>>>>
>>>>>> make[3]: Leaving directory
>>>>>> `/ibrix/home/qcteam/METv4.1/src/libcode/vx_grid'
>>>>>> make[3]: Entering directory
>>>>>> `/ibrix/home/qcteam/METv4.1/src/libcode/vx_nc_util'
>>>>>> /usr/bin/g++ -m32 nc_var_info.cc -Wall -Wshadow -static -g -c
>>>>>> -I/home/qcteam/METv4.1/include
>>>>>> -I/home/qcteam/netcdf/compiled/include
>>>>>> In file included from nc_var_info.cc:24:
>>>>>> nc_var_info.h:24:21: error: netcdf.hh: No such file or
directory
>>>>>>
>>>>>> The nectdf 'include' I compiled for 4.3.0 has only one file in
it,
>>>>>> netcdf.h, while 4.1.3 has a few more, including the netcdf.hh
that
>>>>>> adds the inclusion of one more file from the 4.1.3 include
>>>>>> (netcdfcpp.h).
>>>>>>
>>>>>> The routine say netcdf compiled successfully for me, but I had
to
>>>>>> disable a couple of things to get it to run correctly: (I ran
the
>>>>>> following: "./configure --prefix=/home/qcteam/netcdf/compiled
>>>>>> --disable-netcdf-4 --disable-dap"). Is it possible one of
those
>>>>>> disables is the reason for the missing file(s)?
>>>>>>
>>>>>> It looks as if the timing of the 4.1 release might be using an
>>>>>> older netcdf than 4.3.0? Rather than try every iteration of
>>>>>> netcdf between 4.1.3 and 4.3.0, I thought I'd ask. I wasn't
going
>>>>>> to update netcdf but the compile of 'pb2nc' while using 4.1.3
>>>>>> failed, suggesting I had to update.
>>>>>>
>>>>>> Thanks,
>>>>>> Matt
>>>>>>
>>>>>> // SIGNED //
>>>>>> Matthew C. Sittel
>>>>>> Associate Scientist
>>>>>> University Corporation for Atmospheric Research 16WS/WXN,
Offutt
>>>>>> AFB, NE
>>>>>> (402) 294-3473 DSN: 271-3473
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>
>
>
>
------------------------------------------------
More information about the Met_help
mailing list