[Met_help] [rt.rap.ucar.edu #71608] History for fail to compile MET5.0
John Halley Gotway via RT
met_help at ucar.edu
Mon Aug 17 12:13:02 MDT 2015
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Hello,
I was trying to compile MET5.0 using gnu compiler.
I used gcc version 4.7.3, netcdf version 3.6.0 and gsl version 1.14.
Below is my environment setup:
setenv MET_NETCDF /nasa/netcdf/3.6.0/gcc
setenv MET_BUFRLIB /nobackupp8/yliu15/MET/BUFRLIB
setenv MET_GSL /nasa/gsl/1.14/
When I compiled the MET5.0, I get the error message as "g++: error: unrecognized option '-R'".
I have tried different version of gun compiler and all got the similar error message.
I asked IT and they told me that it is very likely a bug in configure file or Makefile rather than compiler issue.
I have no idea how to solve this problem. Would you please take a look of four files that I attached with this email and let me know how can I make this work.
Thank you very much!
Karry
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: fail to compile MET5.0
From: John Halley Gotway
Time: Thu Apr 23 12:54:34 2015
Karry,
I see that you're having trouble compiling met-5.0. Thanks for
sending all
those log files. It make it easier to debug.
You're getting a compilation error from the PB2NC tool, which is
written in
C++ but links to some Fortran code. This is an odd error message that
I've
never seen before:
g++: error: unrecognized option '-R'
For some reason, a '-R' flag is being inserted into the compilation
command
for Fortran sources. I've never seen that flag before and obviously
the
g++ compiler doesn't like it. I looked at the configure command you
used
and there's nothing odd in there at all.
For now, let's have you compile with the PB2NC tool turned off to
ensure
there aren't other problems. Once everything else is compiling well,
we
can try to solve this issue.
Try the following:
make clean
./configure --prefix=/nobackupp8/yliu15/MET/met-5.0 --disable-pb2nc
make install >& make_install.log
Please let me know how that goes.
Thanks,
John Halley Gotway
met_help at ucar.edu
On Thu, Apr 23, 2015 at 11:27 AM, Liu, Karry at ARB via RT
<met_help at ucar.edu>
wrote:
>
> Thu Apr 23 11:27:44 2015: Request 71608 was acted upon.
> Transaction: Ticket created by Karry.Liu at arb.ca.gov
> Queue: met_help
> Subject: fail to compile MET5.0
> Owner: Nobody
> Requestors: Karry.Liu at arb.ca.gov
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
>
>
> Hello,
> I was trying to compile MET5.0 using gnu compiler.
> I used gcc version 4.7.3, netcdf version 3.6.0 and gsl version 1.14.
> Below is my environment setup:
> setenv MET_NETCDF /nasa/netcdf/3.6.0/gcc
> setenv MET_BUFRLIB /nobackupp8/yliu15/MET/BUFRLIB
> setenv MET_GSL /nasa/gsl/1.14/
>
> When I compiled the MET5.0, I get the error message as "g++: error:
> unrecognized option '-R'".
> I have tried different version of gun compiler and all got the
similar
> error message.
> I asked IT and they told me that it is very likely a bug in
configure file
> or Makefile rather than compiler issue.
> I have no idea how to solve this problem. Would you please take a
look of
> four files that I attached with this email and let me know how can I
make
> this work.
> Thank you very much!
>
> Karry
>
>
>
>
------------------------------------------------
Subject: fail to compile MET5.0
From: Liu, Karry at ARB
Time: Thu Apr 23 14:17:45 2015
Hello John,
I have recompiled the MET 5.0 and get the same error message.
I have attached the log file with this email. Thank you~
Regards,
Karry
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Thursday, April 23, 2015 11:55 AM
To: Liu, Karry at ARB
Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
Karry,
I see that you're having trouble compiling met-5.0. Thanks for
sending all those log files. It make it easier to debug.
You're getting a compilation error from the PB2NC tool, which is
written in
C++ but links to some Fortran code. This is an odd error message that
C++ I've
never seen before:
g++: error: unrecognized option '-R'
For some reason, a '-R' flag is being inserted into the compilation
command for Fortran sources. I've never seen that flag before and
obviously the
g++ compiler doesn't like it. I looked at the configure command you
g++ used
and there's nothing odd in there at all.
For now, let's have you compile with the PB2NC tool turned off to
ensure there aren't other problems. Once everything else is compiling
well, we can try to solve this issue.
Try the following:
make clean
./configure --prefix=/nobackupp8/yliu15/MET/met-5.0 --disable-pb2nc
make install >& make_install.log
Please let me know how that goes.
Thanks,
John Halley Gotway
met_help at ucar.edu
On Thu, Apr 23, 2015 at 11:27 AM, Liu, Karry at ARB via RT
<met_help at ucar.edu>
wrote:
>
> Thu Apr 23 11:27:44 2015: Request 71608 was acted upon.
> Transaction: Ticket created by Karry.Liu at arb.ca.gov
> Queue: met_help
> Subject: fail to compile MET5.0
> Owner: Nobody
> Requestors: Karry.Liu at arb.ca.gov
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608
> >
>
>
> Hello,
> I was trying to compile MET5.0 using gnu compiler.
> I used gcc version 4.7.3, netcdf version 3.6.0 and gsl version 1.14.
> Below is my environment setup:
> setenv MET_NETCDF /nasa/netcdf/3.6.0/gcc setenv MET_BUFRLIB
> /nobackupp8/yliu15/MET/BUFRLIB setenv MET_GSL /nasa/gsl/1.14/
>
> When I compiled the MET5.0, I get the error message as "g++: error:
> unrecognized option '-R'".
> I have tried different version of gun compiler and all got the
similar
> error message.
> I asked IT and they told me that it is very likely a bug in
configure
> file or Makefile rather than compiler issue.
> I have no idea how to solve this problem. Would you please take a
look
> of four files that I attached with this email and let me know how
can
> I make this work.
> Thank you very much!
>
> Karry
>
>
>
>
------------------------------------------------
Subject: fail to compile MET5.0
From: John Halley Gotway
Time: Thu Apr 23 16:22:39 2015
Karry,
Ah, yes. The same issue is popping when compiling the MET Tropical
Cyclone
tools which also link to some Fortran code. Those can be disabled
using
"--disable-tc_utils"
I suspect the following commands would get you a clean compile that
skips
over the PB2NC and MET-TC tools:
make clean
./configure --prefix=/nobackupp8/yliu15/MET/met-5.0 --disable-pb2nc
--disable-tc_utils
make install >& make_install.log
The question remains, how is a "-R" flag making its way into the FLIBS
variable from autoconf? I don't know yet, but will keep digging.
Thanks,
John
On Thu, Apr 23, 2015 at 2:17 PM, Liu, Karry at ARB via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
>
> Hello John,
> I have recompiled the MET 5.0 and get the same error message.
> I have attached the log file with this email. Thank you~
>
> Regards,
>
> Karry
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Thursday, April 23, 2015 11:55 AM
> To: Liu, Karry at ARB
> Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
>
> Karry,
>
> I see that you're having trouble compiling met-5.0. Thanks for
sending
> all those log files. It make it easier to debug.
>
> You're getting a compilation error from the PB2NC tool, which is
written in
> C++ but links to some Fortran code. This is an odd error message
that
> C++ I've
> never seen before:
> g++: error: unrecognized option '-R'
>
> For some reason, a '-R' flag is being inserted into the compilation
> command for Fortran sources. I've never seen that flag before and
> obviously the
> g++ compiler doesn't like it. I looked at the configure command you
> g++ used
> and there's nothing odd in there at all.
>
> For now, let's have you compile with the PB2NC tool turned off to
ensure
> there aren't other problems. Once everything else is compiling
well, we
> can try to solve this issue.
>
> Try the following:
> make clean
> ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0 --disable-
pb2nc
> make install >& make_install.log
>
> Please let me know how that goes.
>
> Thanks,
> John Halley Gotway
> met_help at ucar.edu
>
>
>
> On Thu, Apr 23, 2015 at 11:27 AM, Liu, Karry at ARB via RT
<met_help at ucar.edu
> >
> wrote:
>
> >
> > Thu Apr 23 11:27:44 2015: Request 71608 was acted upon.
> > Transaction: Ticket created by Karry.Liu at arb.ca.gov
> > Queue: met_help
> > Subject: fail to compile MET5.0
> > Owner: Nobody
> > Requestors: Karry.Liu at arb.ca.gov
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608
> > >
> >
> >
> > Hello,
> > I was trying to compile MET5.0 using gnu compiler.
> > I used gcc version 4.7.3, netcdf version 3.6.0 and gsl version
1.14.
> > Below is my environment setup:
> > setenv MET_NETCDF /nasa/netcdf/3.6.0/gcc setenv MET_BUFRLIB
> > /nobackupp8/yliu15/MET/BUFRLIB setenv MET_GSL /nasa/gsl/1.14/
> >
> > When I compiled the MET5.0, I get the error message as "g++:
error:
> > unrecognized option '-R'".
> > I have tried different version of gun compiler and all got the
similar
> > error message.
> > I asked IT and they told me that it is very likely a bug in
configure
> > file or Makefile rather than compiler issue.
> > I have no idea how to solve this problem. Would you please take a
look
> > of four files that I attached with this email and let me know how
can
> > I make this work.
> > Thank you very much!
> >
> > Karry
> >
> >
> >
> >
>
>
>
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #71608] fail to compile MET5.0
From: Liu, Karry at ARB
Time: Thu Apr 23 16:31:26 2015
Hello John,
I just tried this comment, and it seems compile successfully.
To make sure that it really compiled successfully, how many
executable files I should expect under bin?
Karry
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Thursday, April 23, 2015 3:23 PM
To: Liu, Karry at ARB
Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
Karry,
Ah, yes. The same issue is popping when compiling the MET Tropical
Cyclone tools which also link to some Fortran code. Those can be
disabled using "--disable-tc_utils"
I suspect the following commands would get you a clean compile that
skips over the PB2NC and MET-TC tools:
make clean
./configure --prefix=/nobackupp8/yliu15/MET/met-5.0 --disable-pb2nc
--disable-tc_utils
make install >& make_install.log
The question remains, how is a "-R" flag making its way into the FLIBS
variable from autoconf? I don't know yet, but will keep digging.
Thanks,
John
On Thu, Apr 23, 2015 at 2:17 PM, Liu, Karry at ARB via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
>
> Hello John,
> I have recompiled the MET 5.0 and get the same error message.
> I have attached the log file with this email. Thank you~
>
> Regards,
>
> Karry
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Thursday, April 23, 2015 11:55 AM
> To: Liu, Karry at ARB
> Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
>
> Karry,
>
> I see that you're having trouble compiling met-5.0. Thanks for
> sending all those log files. It make it easier to debug.
>
> You're getting a compilation error from the PB2NC tool, which is
> written in
> C++ but links to some Fortran code. This is an odd error message
that
> C++ I've
> never seen before:
> g++: error: unrecognized option '-R'
>
> For some reason, a '-R' flag is being inserted into the compilation
> command for Fortran sources. I've never seen that flag before and
> obviously the
> g++ compiler doesn't like it. I looked at the configure command you
> g++ used
> and there's nothing odd in there at all.
>
> For now, let's have you compile with the PB2NC tool turned off to
> ensure there aren't other problems. Once everything else is
compiling
> well, we can try to solve this issue.
>
> Try the following:
> make clean
> ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0 --disable-
pb2nc
> make install >& make_install.log
>
> Please let me know how that goes.
>
> Thanks,
> John Halley Gotway
> met_help at ucar.edu
>
>
>
> On Thu, Apr 23, 2015 at 11:27 AM, Liu, Karry at ARB via RT
> <met_help at ucar.edu
> >
> wrote:
>
> >
> > Thu Apr 23 11:27:44 2015: Request 71608 was acted upon.
> > Transaction: Ticket created by Karry.Liu at arb.ca.gov
> > Queue: met_help
> > Subject: fail to compile MET5.0
> > Owner: Nobody
> > Requestors: Karry.Liu at arb.ca.gov
> > Status: new
> > Ticket <URL:
> > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608
> > >
> >
> >
> > Hello,
> > I was trying to compile MET5.0 using gnu compiler.
> > I used gcc version 4.7.3, netcdf version 3.6.0 and gsl version
1.14.
> > Below is my environment setup:
> > setenv MET_NETCDF /nasa/netcdf/3.6.0/gcc setenv MET_BUFRLIB
> > /nobackupp8/yliu15/MET/BUFRLIB setenv MET_GSL /nasa/gsl/1.14/
> >
> > When I compiled the MET5.0, I get the error message as "g++:
error:
> > unrecognized option '-R'".
> > I have tried different version of gun compiler and all got the
> > similar error message.
> > I asked IT and they told me that it is very likely a bug in
> > configure file or Makefile rather than compiler issue.
> > I have no idea how to solve this problem. Would you please take a
> > look of four files that I attached with this email and let me know
> > how can I make this work.
> > Thank you very much!
> >
> > Karry
> >
> >
> >
> >
>
>
>
------------------------------------------------
Subject: fail to compile MET5.0
From: John Halley Gotway
Time: Thu Apr 23 16:37:43 2015
Karry,
I'd expect 23 minus pb2nc, tc_dland, tc_pairs, tc_stat, modis_regrid,
and
plot_mode_field which equals 17:
ascii2nc*
ensemble_stat*
gen_circle_mask*
gen_poly_mask*
grid_stat*
madis2nc*
mode*
mode_analysis*
pcp_combine*
plot_data_plane*
plot_point_obs*
point_stat*
series_analysis*
stat_analysis*
wavelet_stat*
wwmca_plot*
wwmca_regrid*
Is that what you're seeing?
Thanks,
John
On Thu, Apr 23, 2015 at 4:31 PM, Liu, Karry at ARB via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
>
> Hello John,
> I just tried this comment, and it seems compile successfully.
> To make sure that it really compiled successfully, how many
executable
> files I should expect under bin?
>
> Karry
>
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Thursday, April 23, 2015 3:23 PM
> To: Liu, Karry at ARB
> Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
>
> Karry,
>
> Ah, yes. The same issue is popping when compiling the MET Tropical
> Cyclone tools which also link to some Fortran code. Those can be
disabled
> using "--disable-tc_utils"
>
> I suspect the following commands would get you a clean compile that
skips
> over the PB2NC and MET-TC tools:
> make clean
> ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0 --disable-
pb2nc
> --disable-tc_utils
> make install >& make_install.log
>
> The question remains, how is a "-R" flag making its way into the
FLIBS
> variable from autoconf? I don't know yet, but will keep digging.
>
> Thanks,
> John
>
> On Thu, Apr 23, 2015 at 2:17 PM, Liu, Karry at ARB via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
> >
> > Hello John,
> > I have recompiled the MET 5.0 and get the same error message.
> > I have attached the log file with this email. Thank you~
> >
> > Regards,
> >
> > Karry
> >
> > -----Original Message-----
> > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > Sent: Thursday, April 23, 2015 11:55 AM
> > To: Liu, Karry at ARB
> > Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
> >
> > Karry,
> >
> > I see that you're having trouble compiling met-5.0. Thanks for
> > sending all those log files. It make it easier to debug.
> >
> > You're getting a compilation error from the PB2NC tool, which is
> > written in
> > C++ but links to some Fortran code. This is an odd error message
that
> > C++ I've
> > never seen before:
> > g++: error: unrecognized option '-R'
> >
> > For some reason, a '-R' flag is being inserted into the
compilation
> > command for Fortran sources. I've never seen that flag before and
> > obviously the
> > g++ compiler doesn't like it. I looked at the configure command
you
> > g++ used
> > and there's nothing odd in there at all.
> >
> > For now, let's have you compile with the PB2NC tool turned off to
> > ensure there aren't other problems. Once everything else is
compiling
> > well, we can try to solve this issue.
> >
> > Try the following:
> > make clean
> > ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0 --disable-
pb2nc
> > make install >& make_install.log
> >
> > Please let me know how that goes.
> >
> > Thanks,
> > John Halley Gotway
> > met_help at ucar.edu
> >
> >
> >
> > On Thu, Apr 23, 2015 at 11:27 AM, Liu, Karry at ARB via RT
> > <met_help at ucar.edu
> > >
> > wrote:
> >
> > >
> > > Thu Apr 23 11:27:44 2015: Request 71608 was acted upon.
> > > Transaction: Ticket created by Karry.Liu at arb.ca.gov
> > > Queue: met_help
> > > Subject: fail to compile MET5.0
> > > Owner: Nobody
> > > Requestors: Karry.Liu at arb.ca.gov
> > > Status: new
> > > Ticket <URL:
> > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608
> > > >
> > >
> > >
> > > Hello,
> > > I was trying to compile MET5.0 using gnu compiler.
> > > I used gcc version 4.7.3, netcdf version 3.6.0 and gsl version
1.14.
> > > Below is my environment setup:
> > > setenv MET_NETCDF /nasa/netcdf/3.6.0/gcc setenv MET_BUFRLIB
> > > /nobackupp8/yliu15/MET/BUFRLIB setenv MET_GSL /nasa/gsl/1.14/
> > >
> > > When I compiled the MET5.0, I get the error message as "g++:
error:
> > > unrecognized option '-R'".
> > > I have tried different version of gun compiler and all got the
> > > similar error message.
> > > I asked IT and they told me that it is very likely a bug in
> > > configure file or Makefile rather than compiler issue.
> > > I have no idea how to solve this problem. Would you please take
a
> > > look of four files that I attached with this email and let me
know
> > > how can I make this work.
> > > Thank you very much!
> > >
> > > Karry
> > >
> > >
> > >
> > >
> >
> >
> >
>
>
>
>
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #71608] fail to compile MET5.0
From: Liu, Karry at ARB
Time: Thu Apr 23 17:05:27 2015
Hello John,
Yes, I got 17 of these files.
Karry
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Thursday, April 23, 2015 3:38 PM
To: Liu, Karry at ARB
Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
Karry,
I'd expect 23 minus pb2nc, tc_dland, tc_pairs, tc_stat, modis_regrid,
and plot_mode_field which equals 17:
ascii2nc*
ensemble_stat*
gen_circle_mask*
gen_poly_mask*
grid_stat*
madis2nc*
mode*
mode_analysis*
pcp_combine*
plot_data_plane*
plot_point_obs*
point_stat*
series_analysis*
stat_analysis*
wavelet_stat*
wwmca_plot*
wwmca_regrid*
Is that what you're seeing?
Thanks,
John
On Thu, Apr 23, 2015 at 4:31 PM, Liu, Karry at ARB via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
>
> Hello John,
> I just tried this comment, and it seems compile successfully.
> To make sure that it really compiled successfully, how many
> executable files I should expect under bin?
>
> Karry
>
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Thursday, April 23, 2015 3:23 PM
> To: Liu, Karry at ARB
> Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
>
> Karry,
>
> Ah, yes. The same issue is popping when compiling the MET Tropical
> Cyclone tools which also link to some Fortran code. Those can be
> disabled using "--disable-tc_utils"
>
> I suspect the following commands would get you a clean compile that
> skips over the PB2NC and MET-TC tools:
> make clean
> ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0 --disable-
pb2nc
> --disable-tc_utils
> make install >& make_install.log
>
> The question remains, how is a "-R" flag making its way into the
FLIBS
> variable from autoconf? I don't know yet, but will keep digging.
>
> Thanks,
> John
>
> On Thu, Apr 23, 2015 at 2:17 PM, Liu, Karry at ARB via RT
> <met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
> >
> > Hello John,
> > I have recompiled the MET 5.0 and get the same error message.
> > I have attached the log file with this email. Thank you~
> >
> > Regards,
> >
> > Karry
> >
> > -----Original Message-----
> > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > Sent: Thursday, April 23, 2015 11:55 AM
> > To: Liu, Karry at ARB
> > Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
> >
> > Karry,
> >
> > I see that you're having trouble compiling met-5.0. Thanks for
> > sending all those log files. It make it easier to debug.
> >
> > You're getting a compilation error from the PB2NC tool, which is
> > written in
> > C++ but links to some Fortran code. This is an odd error message
> > C++ that I've
> > never seen before:
> > g++: error: unrecognized option '-R'
> >
> > For some reason, a '-R' flag is being inserted into the
compilation
> > command for Fortran sources. I've never seen that flag before and
> > obviously the
> > g++ compiler doesn't like it. I looked at the configure command
you
> > g++ used
> > and there's nothing odd in there at all.
> >
> > For now, let's have you compile with the PB2NC tool turned off to
> > ensure there aren't other problems. Once everything else is
> > compiling well, we can try to solve this issue.
> >
> > Try the following:
> > make clean
> > ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0 --disable-
pb2nc
> > make install >& make_install.log
> >
> > Please let me know how that goes.
> >
> > Thanks,
> > John Halley Gotway
> > met_help at ucar.edu
> >
> >
> >
> > On Thu, Apr 23, 2015 at 11:27 AM, Liu, Karry at ARB via RT
> > <met_help at ucar.edu
> > >
> > wrote:
> >
> > >
> > > Thu Apr 23 11:27:44 2015: Request 71608 was acted upon.
> > > Transaction: Ticket created by Karry.Liu at arb.ca.gov
> > > Queue: met_help
> > > Subject: fail to compile MET5.0
> > > Owner: Nobody
> > > Requestors: Karry.Liu at arb.ca.gov
> > > Status: new
> > > Ticket <URL:
> > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608
> > > >
> > >
> > >
> > > Hello,
> > > I was trying to compile MET5.0 using gnu compiler.
> > > I used gcc version 4.7.3, netcdf version 3.6.0 and gsl version
1.14.
> > > Below is my environment setup:
> > > setenv MET_NETCDF /nasa/netcdf/3.6.0/gcc setenv MET_BUFRLIB
> > > /nobackupp8/yliu15/MET/BUFRLIB setenv MET_GSL /nasa/gsl/1.14/
> > >
> > > When I compiled the MET5.0, I get the error message as "g++:
error:
> > > unrecognized option '-R'".
> > > I have tried different version of gun compiler and all got the
> > > similar error message.
> > > I asked IT and they told me that it is very likely a bug in
> > > configure file or Makefile rather than compiler issue.
> > > I have no idea how to solve this problem. Would you please take
a
> > > look of four files that I attached with this email and let me
know
> > > how can I make this work.
> > > Thank you very much!
> > >
> > > Karry
> > >
> > >
> > >
> > >
> >
> >
> >
>
>
>
>
------------------------------------------------
Subject: fail to compile MET5.0
From: John Halley Gotway
Time: Fri Apr 24 10:21:27 2015
Karry,
Great, glad you were able to compile the other tools. Still would be
nice
to resolve this issue of getting the C++ and Fortran code to
communicate.
I've searched around and can find no mention of a "-R" option. So I
have
no idea how it's making it into the output of the configure command.
One possibility is that "-R" is defined somewhere in your environment
and
its getting passed through by configure. Try running this command:
printenv | grep "\-R"
Wherever you find "-R", edit your shell configuration file and remove
it.
And here's another work-around. You could try manually setting FLIBS
before running configure. In this settings, I've just switched "-R"
to
"-L".
setenv FLIBS -L${MET_NETCDF}/lib -L${MET_GSL}/lib
-L/nobackupp8/yliu15/MET/BUFRLIB
-L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-linux/4.7.3
-L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../../../x86_64-unknown-linux/lib/../lib64
-L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../../../lib64
-L/lib/../lib64 -L/usr/lib/../lib64 -L/nasa/gsl/1.14/lib
-L/nasa/netcdf/3.6.0/gcc/lib
-L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../../../x86_64-unknown-linux/lib
-L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../..
-L/nasa/pkgsrc/2014Q2/gcc47/x86_64-unknown-linux/lib/../lib64
-L/nasa/pkgsrc/2014Q2/gcc47/lib/../lib64 -lgfortran -lm -lquadmath
make clean
./configure --prefix=/nobackupp8/yliu15/MET/met-5.0
make install >& make_install.log
Please let me know how it goes.
Thanks,
John
On Thu, Apr 23, 2015 at 5:05 PM, Liu, Karry at ARB via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
>
> Hello John,
> Yes, I got 17 of these files.
>
> Karry
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Thursday, April 23, 2015 3:38 PM
> To: Liu, Karry at ARB
> Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
>
> Karry,
>
> I'd expect 23 minus pb2nc, tc_dland, tc_pairs, tc_stat,
modis_regrid, and
> plot_mode_field which equals 17:
> ascii2nc*
> ensemble_stat*
> gen_circle_mask*
> gen_poly_mask*
> grid_stat*
> madis2nc*
> mode*
> mode_analysis*
> pcp_combine*
> plot_data_plane*
> plot_point_obs*
> point_stat*
> series_analysis*
> stat_analysis*
> wavelet_stat*
> wwmca_plot*
> wwmca_regrid*
>
> Is that what you're seeing?
>
> Thanks,
> John
>
>
> On Thu, Apr 23, 2015 at 4:31 PM, Liu, Karry at ARB via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
> >
> > Hello John,
> > I just tried this comment, and it seems compile successfully.
> > To make sure that it really compiled successfully, how many
> > executable files I should expect under bin?
> >
> > Karry
> >
> >
> > -----Original Message-----
> > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > Sent: Thursday, April 23, 2015 3:23 PM
> > To: Liu, Karry at ARB
> > Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
> >
> > Karry,
> >
> > Ah, yes. The same issue is popping when compiling the MET
Tropical
> > Cyclone tools which also link to some Fortran code. Those can be
> > disabled using "--disable-tc_utils"
> >
> > I suspect the following commands would get you a clean compile
that
> > skips over the PB2NC and MET-TC tools:
> > make clean
> > ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0 --disable-
pb2nc
> > --disable-tc_utils
> > make install >& make_install.log
> >
> > The question remains, how is a "-R" flag making its way into the
FLIBS
> > variable from autoconf? I don't know yet, but will keep digging.
> >
> > Thanks,
> > John
> >
> > On Thu, Apr 23, 2015 at 2:17 PM, Liu, Karry at ARB via RT
> > <met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
> > >
> > > Hello John,
> > > I have recompiled the MET 5.0 and get the same error message.
> > > I have attached the log file with this email. Thank you~
> > >
> > > Regards,
> > >
> > > Karry
> > >
> > > -----Original Message-----
> > > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > > Sent: Thursday, April 23, 2015 11:55 AM
> > > To: Liu, Karry at ARB
> > > Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
> > >
> > > Karry,
> > >
> > > I see that you're having trouble compiling met-5.0. Thanks for
> > > sending all those log files. It make it easier to debug.
> > >
> > > You're getting a compilation error from the PB2NC tool, which is
> > > written in
> > > C++ but links to some Fortran code. This is an odd error
message
> > > C++ that I've
> > > never seen before:
> > > g++: error: unrecognized option '-R'
> > >
> > > For some reason, a '-R' flag is being inserted into the
compilation
> > > command for Fortran sources. I've never seen that flag before
and
> > > obviously the
> > > g++ compiler doesn't like it. I looked at the configure command
you
> > > g++ used
> > > and there's nothing odd in there at all.
> > >
> > > For now, let's have you compile with the PB2NC tool turned off
to
> > > ensure there aren't other problems. Once everything else is
> > > compiling well, we can try to solve this issue.
> > >
> > > Try the following:
> > > make clean
> > > ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0
--disable-pb2nc
> > > make install >& make_install.log
> > >
> > > Please let me know how that goes.
> > >
> > > Thanks,
> > > John Halley Gotway
> > > met_help at ucar.edu
> > >
> > >
> > >
> > > On Thu, Apr 23, 2015 at 11:27 AM, Liu, Karry at ARB via RT
> > > <met_help at ucar.edu
> > > >
> > > wrote:
> > >
> > > >
> > > > Thu Apr 23 11:27:44 2015: Request 71608 was acted upon.
> > > > Transaction: Ticket created by Karry.Liu at arb.ca.gov
> > > > Queue: met_help
> > > > Subject: fail to compile MET5.0
> > > > Owner: Nobody
> > > > Requestors: Karry.Liu at arb.ca.gov
> > > > Status: new
> > > > Ticket <URL:
> > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608
> > > > >
> > > >
> > > >
> > > > Hello,
> > > > I was trying to compile MET5.0 using gnu compiler.
> > > > I used gcc version 4.7.3, netcdf version 3.6.0 and gsl version
1.14.
> > > > Below is my environment setup:
> > > > setenv MET_NETCDF /nasa/netcdf/3.6.0/gcc setenv MET_BUFRLIB
> > > > /nobackupp8/yliu15/MET/BUFRLIB setenv MET_GSL /nasa/gsl/1.14/
> > > >
> > > > When I compiled the MET5.0, I get the error message as "g++:
error:
> > > > unrecognized option '-R'".
> > > > I have tried different version of gun compiler and all got the
> > > > similar error message.
> > > > I asked IT and they told me that it is very likely a bug in
> > > > configure file or Makefile rather than compiler issue.
> > > > I have no idea how to solve this problem. Would you please
take a
> > > > look of four files that I attached with this email and let me
know
> > > > how can I make this work.
> > > > Thank you very much!
> > > >
> > > > Karry
> > > >
> > > >
> > > >
> > > >
> > >
> > >
> > >
> >
> >
> >
> >
>
>
>
>
------------------------------------------------
Subject: fail to compile MET5.0
From: Liu, Karry at ARB
Time: Fri Apr 24 11:36:55 2015
Hello John,
I tried the first one, but it does not work since there is no -R in
my environment setup.
I tried the second one, and it works! I got 21 executable files.
However, when I used make test, and there are some errors in the log
file.
I am not sure if those errors matter? Would you take a look of log
file that I attached with this email?
Thank you very much for all the help!
Karry
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Friday, April 24, 2015 9:21 AM
To: Liu, Karry at ARB
Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
Karry,
Great, glad you were able to compile the other tools. Still would be
nice to resolve this issue of getting the C++ and Fortran code to
communicate.
I've searched around and can find no mention of a "-R" option. So I
have no idea how it's making it into the output of the configure
command.
One possibility is that "-R" is defined somewhere in your environment
and its getting passed through by configure. Try running this
command:
printenv | grep "\-R"
Wherever you find "-R", edit your shell configuration file and remove
it.
And here's another work-around. You could try manually setting FLIBS
before running configure. In this settings, I've just switched "-R"
to "-L".
setenv FLIBS -L${MET_NETCDF}/lib -L${MET_GSL}/lib
-L/nobackupp8/yliu15/MET/BUFRLIB
-L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-linux/4.7.3
-L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../../../x86_64-unknown-linux/lib/../lib64
-L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../../../lib64
-L/lib/../lib64 -L/usr/lib/../lib64 -L/nasa/gsl/1.14/lib
-L/nasa/netcdf/3.6.0/gcc/lib
-L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../../../x86_64-unknown-linux/lib
-L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../..
-L/nasa/pkgsrc/2014Q2/gcc47/x86_64-unknown-linux/lib/../lib64
-L/nasa/pkgsrc/2014Q2/gcc47/lib/../lib64 -lgfortran -lm -lquadmath
make clean
./configure --prefix=/nobackupp8/yliu15/MET/met-5.0
make install >& make_install.log
Please let me know how it goes.
Thanks,
John
On Thu, Apr 23, 2015 at 5:05 PM, Liu, Karry at ARB via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
>
> Hello John,
> Yes, I got 17 of these files.
>
> Karry
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Thursday, April 23, 2015 3:38 PM
> To: Liu, Karry at ARB
> Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
>
> Karry,
>
> I'd expect 23 minus pb2nc, tc_dland, tc_pairs, tc_stat,
modis_regrid,
> and plot_mode_field which equals 17:
> ascii2nc*
> ensemble_stat*
> gen_circle_mask*
> gen_poly_mask*
> grid_stat*
> madis2nc*
> mode*
> mode_analysis*
> pcp_combine*
> plot_data_plane*
> plot_point_obs*
> point_stat*
> series_analysis*
> stat_analysis*
> wavelet_stat*
> wwmca_plot*
> wwmca_regrid*
>
> Is that what you're seeing?
>
> Thanks,
> John
>
>
> On Thu, Apr 23, 2015 at 4:31 PM, Liu, Karry at ARB via RT
> <met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
> >
> > Hello John,
> > I just tried this comment, and it seems compile successfully.
> > To make sure that it really compiled successfully, how many
> > executable files I should expect under bin?
> >
> > Karry
> >
> >
> > -----Original Message-----
> > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > Sent: Thursday, April 23, 2015 3:23 PM
> > To: Liu, Karry at ARB
> > Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
> >
> > Karry,
> >
> > Ah, yes. The same issue is popping when compiling the MET
Tropical
> > Cyclone tools which also link to some Fortran code. Those can be
> > disabled using "--disable-tc_utils"
> >
> > I suspect the following commands would get you a clean compile
that
> > skips over the PB2NC and MET-TC tools:
> > make clean
> > ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0
> > --disable-pb2nc --disable-tc_utils
> > make install >& make_install.log
> >
> > The question remains, how is a "-R" flag making its way into the
> > FLIBS variable from autoconf? I don't know yet, but will keep
digging.
> >
> > Thanks,
> > John
> >
> > On Thu, Apr 23, 2015 at 2:17 PM, Liu, Karry at ARB via RT
> > <met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
> > >
> > > Hello John,
> > > I have recompiled the MET 5.0 and get the same error message.
> > > I have attached the log file with this email. Thank you~
> > >
> > > Regards,
> > >
> > > Karry
> > >
> > > -----Original Message-----
> > > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > > Sent: Thursday, April 23, 2015 11:55 AM
> > > To: Liu, Karry at ARB
> > > Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
> > >
> > > Karry,
> > >
> > > I see that you're having trouble compiling met-5.0. Thanks for
> > > sending all those log files. It make it easier to debug.
> > >
> > > You're getting a compilation error from the PB2NC tool, which is
> > > written in
> > > C++ but links to some Fortran code. This is an odd error
message
> > > C++ that I've
> > > never seen before:
> > > g++: error: unrecognized option '-R'
> > >
> > > For some reason, a '-R' flag is being inserted into the
> > > compilation command for Fortran sources. I've never seen that
> > > flag before and obviously the
> > > g++ compiler doesn't like it. I looked at the configure command
> > > g++ you used
> > > and there's nothing odd in there at all.
> > >
> > > For now, let's have you compile with the PB2NC tool turned off
to
> > > ensure there aren't other problems. Once everything else is
> > > compiling well, we can try to solve this issue.
> > >
> > > Try the following:
> > > make clean
> > > ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0
--disable-pb2nc
> > > make install >& make_install.log
> > >
> > > Please let me know how that goes.
> > >
> > > Thanks,
> > > John Halley Gotway
> > > met_help at ucar.edu
> > >
> > >
> > >
> > > On Thu, Apr 23, 2015 at 11:27 AM, Liu, Karry at ARB via RT
> > > <met_help at ucar.edu
> > > >
> > > wrote:
> > >
> > > >
> > > > Thu Apr 23 11:27:44 2015: Request 71608 was acted upon.
> > > > Transaction: Ticket created by Karry.Liu at arb.ca.gov
> > > > Queue: met_help
> > > > Subject: fail to compile MET5.0
> > > > Owner: Nobody
> > > > Requestors: Karry.Liu at arb.ca.gov
> > > > Status: new
> > > > Ticket <URL:
> > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608
> > > > >
> > > >
> > > >
> > > > Hello,
> > > > I was trying to compile MET5.0 using gnu compiler.
> > > > I used gcc version 4.7.3, netcdf version 3.6.0 and gsl version
1.14.
> > > > Below is my environment setup:
> > > > setenv MET_NETCDF /nasa/netcdf/3.6.0/gcc setenv MET_BUFRLIB
> > > > /nobackupp8/yliu15/MET/BUFRLIB setenv MET_GSL /nasa/gsl/1.14/
> > > >
> > > > When I compiled the MET5.0, I get the error message as "g++:
error:
> > > > unrecognized option '-R'".
> > > > I have tried different version of gun compiler and all got the
> > > > similar error message.
> > > > I asked IT and they told me that it is very likely a bug in
> > > > configure file or Makefile rather than compiler issue.
> > > > I have no idea how to solve this problem. Would you please
take
> > > > a look of four files that I attached with this email and let
me
> > > > know how can I make this work.
> > > > Thank you very much!
> > > >
> > > > Karry
> > > >
> > > >
> > > >
> > > >
> > >
> > >
> > >
> >
> >
> >
> >
>
>
>
>
------------------------------------------------
Subject: fail to compile MET5.0
From: John Halley Gotway
Time: Fri Apr 24 12:29:42 2015
Karry,
Thanks for sending the test output log file.
Since the output of the test scripts rely on each other, the errors at
the
beginning of the tests are causing the errors further down.
I see that the gen_poly_mask tool is successfully writing a NetCDF
file,
but the pcp_combine tool errors out when trying to write a NetCDF
file:
ERROR : write_netcdf() -> error with pcp_var->put()
The question is why? Let's check to see if the call to gen_poly_mask
actually did create good output. Try running the following command
from
the top-level met-5.0 directory:
bin/plot_data_plane \
out/gen_poly_mask/CONUS_poly.nc \
out/gen_poly_mask/CONUS_poly.ps \
'name="CONUS"; level="(*,*)";'
And then view that postscript output file, using something like "gv"
for
ghostview:
gv out/gen_poly_mask/CONUS_poly.ps
Do you see a map of 0's and 1's over the USA?
I'm really not sure why we're getting this runtime error. I see that
you've compiling against NetCDF version 3.6.0. I suppose we could try
a
newer version. Not sure if that would help?
Thanks,
John
On Fri, Apr 24, 2015 at 11:36 AM, Liu, Karry at ARB via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
>
> Hello John,
> I tried the first one, but it does not work since there is no -R
in my
> environment setup.
> I tried the second one, and it works! I got 21 executable files.
> However, when I used make test, and there are some errors in the log
file.
> I am not sure if those errors matter? Would you take a look of log
file
> that I attached with this email?
>
> Thank you very much for all the help!
>
> Karry
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Friday, April 24, 2015 9:21 AM
> To: Liu, Karry at ARB
> Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
>
> Karry,
>
> Great, glad you were able to compile the other tools. Still would
be nice
> to resolve this issue of getting the C++ and Fortran code to
communicate.
> I've searched around and can find no mention of a "-R" option. So I
have
> no idea how it's making it into the output of the configure command.
>
> One possibility is that "-R" is defined somewhere in your
environment and
> its getting passed through by configure. Try running this command:
> printenv | grep "\-R"
>
> Wherever you find "-R", edit your shell configuration file and
remove it.
>
> And here's another work-around. You could try manually setting
FLIBS
> before running configure. In this settings, I've just switched "-R"
to
> "-L".
>
> setenv FLIBS -L${MET_NETCDF}/lib -L${MET_GSL}/lib
> -L/nobackupp8/yliu15/MET/BUFRLIB
> -L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-linux/4.7.3
>
> -L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../../../x86_64-unknown-linux/lib/../lib64
>
> -L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../../../lib64
> -L/lib/../lib64 -L/usr/lib/../lib64 -L/nasa/gsl/1.14/lib
> -L/nasa/netcdf/3.6.0/gcc/lib
> -L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../../../x86_64-unknown-linux/lib
> -L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../..
> -L/nasa/pkgsrc/2014Q2/gcc47/x86_64-unknown-linux/lib/../lib64
> -L/nasa/pkgsrc/2014Q2/gcc47/lib/../lib64 -lgfortran -lm -lquadmath
> make clean
> ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0
> make install >& make_install.log
>
> Please let me know how it goes.
>
> Thanks,
> John
>
> On Thu, Apr 23, 2015 at 5:05 PM, Liu, Karry at ARB via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
> >
> > Hello John,
> > Yes, I got 17 of these files.
> >
> > Karry
> >
> > -----Original Message-----
> > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > Sent: Thursday, April 23, 2015 3:38 PM
> > To: Liu, Karry at ARB
> > Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
> >
> > Karry,
> >
> > I'd expect 23 minus pb2nc, tc_dland, tc_pairs, tc_stat,
modis_regrid,
> > and plot_mode_field which equals 17:
> > ascii2nc*
> > ensemble_stat*
> > gen_circle_mask*
> > gen_poly_mask*
> > grid_stat*
> > madis2nc*
> > mode*
> > mode_analysis*
> > pcp_combine*
> > plot_data_plane*
> > plot_point_obs*
> > point_stat*
> > series_analysis*
> > stat_analysis*
> > wavelet_stat*
> > wwmca_plot*
> > wwmca_regrid*
> >
> > Is that what you're seeing?
> >
> > Thanks,
> > John
> >
> >
> > On Thu, Apr 23, 2015 at 4:31 PM, Liu, Karry at ARB via RT
> > <met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
> > >
> > > Hello John,
> > > I just tried this comment, and it seems compile successfully.
> > > To make sure that it really compiled successfully, how many
> > > executable files I should expect under bin?
> > >
> > > Karry
> > >
> > >
> > > -----Original Message-----
> > > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > > Sent: Thursday, April 23, 2015 3:23 PM
> > > To: Liu, Karry at ARB
> > > Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
> > >
> > > Karry,
> > >
> > > Ah, yes. The same issue is popping when compiling the MET
Tropical
> > > Cyclone tools which also link to some Fortran code. Those can
be
> > > disabled using "--disable-tc_utils"
> > >
> > > I suspect the following commands would get you a clean compile
that
> > > skips over the PB2NC and MET-TC tools:
> > > make clean
> > > ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0
> > > --disable-pb2nc --disable-tc_utils
> > > make install >& make_install.log
> > >
> > > The question remains, how is a "-R" flag making its way into the
> > > FLIBS variable from autoconf? I don't know yet, but will keep
digging.
> > >
> > > Thanks,
> > > John
> > >
> > > On Thu, Apr 23, 2015 at 2:17 PM, Liu, Karry at ARB via RT
> > > <met_help at ucar.edu>
> > > wrote:
> > >
> > > >
> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608
>
> > > >
> > > > Hello John,
> > > > I have recompiled the MET 5.0 and get the same error
message.
> > > > I have attached the log file with this email. Thank you~
> > > >
> > > > Regards,
> > > >
> > > > Karry
> > > >
> > > > -----Original Message-----
> > > > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > > > Sent: Thursday, April 23, 2015 11:55 AM
> > > > To: Liu, Karry at ARB
> > > > Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
> > > >
> > > > Karry,
> > > >
> > > > I see that you're having trouble compiling met-5.0. Thanks
for
> > > > sending all those log files. It make it easier to debug.
> > > >
> > > > You're getting a compilation error from the PB2NC tool, which
is
> > > > written in
> > > > C++ but links to some Fortran code. This is an odd error
message
> > > > C++ that I've
> > > > never seen before:
> > > > g++: error: unrecognized option '-R'
> > > >
> > > > For some reason, a '-R' flag is being inserted into the
> > > > compilation command for Fortran sources. I've never seen that
> > > > flag before and obviously the
> > > > g++ compiler doesn't like it. I looked at the configure
command
> > > > g++ you used
> > > > and there's nothing odd in there at all.
> > > >
> > > > For now, let's have you compile with the PB2NC tool turned off
to
> > > > ensure there aren't other problems. Once everything else is
> > > > compiling well, we can try to solve this issue.
> > > >
> > > > Try the following:
> > > > make clean
> > > > ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0
> --disable-pb2nc
> > > > make install >& make_install.log
> > > >
> > > > Please let me know how that goes.
> > > >
> > > > Thanks,
> > > > John Halley Gotway
> > > > met_help at ucar.edu
> > > >
> > > >
> > > >
> > > > On Thu, Apr 23, 2015 at 11:27 AM, Liu, Karry at ARB via RT
> > > > <met_help at ucar.edu
> > > > >
> > > > wrote:
> > > >
> > > > >
> > > > > Thu Apr 23 11:27:44 2015: Request 71608 was acted upon.
> > > > > Transaction: Ticket created by Karry.Liu at arb.ca.gov
> > > > > Queue: met_help
> > > > > Subject: fail to compile MET5.0
> > > > > Owner: Nobody
> > > > > Requestors: Karry.Liu at arb.ca.gov
> > > > > Status: new
> > > > > Ticket <URL:
> > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608
> > > > > >
> > > > >
> > > > >
> > > > > Hello,
> > > > > I was trying to compile MET5.0 using gnu compiler.
> > > > > I used gcc version 4.7.3, netcdf version 3.6.0 and gsl
version
> 1.14.
> > > > > Below is my environment setup:
> > > > > setenv MET_NETCDF /nasa/netcdf/3.6.0/gcc setenv MET_BUFRLIB
> > > > > /nobackupp8/yliu15/MET/BUFRLIB setenv MET_GSL
/nasa/gsl/1.14/
> > > > >
> > > > > When I compiled the MET5.0, I get the error message as
"g++:
> error:
> > > > > unrecognized option '-R'".
> > > > > I have tried different version of gun compiler and all got
the
> > > > > similar error message.
> > > > > I asked IT and they told me that it is very likely a bug in
> > > > > configure file or Makefile rather than compiler issue.
> > > > > I have no idea how to solve this problem. Would you please
take
> > > > > a look of four files that I attached with this email and let
me
> > > > > know how can I make this work.
> > > > > Thank you very much!
> > > > >
> > > > > Karry
> > > > >
> > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > >
> > >
> > >
> > >
> >
> >
> >
> >
>
>
>
------------------------------------------------
Subject: fail to compile MET5.0
From: Liu, Karry at ARB
Time: Fri Apr 24 13:38:56 2015
Hello John,
I used newer version of netcdf4.0, and those error are gone.
However I got the one last error for "Running PB2NC on a PREPBUFR
file"
I tried to use add --disable-block4, but still have this error.
I have attached the log file with this email. Thank you.
Regards,
Karry
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Friday, April 24, 2015 11:30 AM
To: Liu, Karry at ARB
Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
Karry,
Thanks for sending the test output log file.
Since the output of the test scripts rely on each other, the errors at
the beginning of the tests are causing the errors further down.
I see that the gen_poly_mask tool is successfully writing a NetCDF
file, but the pcp_combine tool errors out when trying to write a
NetCDF file:
ERROR : write_netcdf() -> error with pcp_var->put()
The question is why? Let's check to see if the call to gen_poly_mask
actually did create good output. Try running the following command
from the top-level met-5.0 directory:
bin/plot_data_plane \
out/gen_poly_mask/CONUS_poly.nc \
out/gen_poly_mask/CONUS_poly.ps \
'name="CONUS"; level="(*,*)";'
And then view that postscript output file, using something like "gv"
for
ghostview:
gv out/gen_poly_mask/CONUS_poly.ps
Do you see a map of 0's and 1's over the USA?
I'm really not sure why we're getting this runtime error. I see that
you've compiling against NetCDF version 3.6.0. I suppose we could try
a newer version. Not sure if that would help?
Thanks,
John
On Fri, Apr 24, 2015 at 11:36 AM, Liu, Karry at ARB via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
>
> Hello John,
> I tried the first one, but it does not work since there is no -R
in
> my environment setup.
> I tried the second one, and it works! I got 21 executable files.
> However, when I used make test, and there are some errors in the log
file.
> I am not sure if those errors matter? Would you take a look of log
> file that I attached with this email?
>
> Thank you very much for all the help!
>
> Karry
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Friday, April 24, 2015 9:21 AM
> To: Liu, Karry at ARB
> Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
>
> Karry,
>
> Great, glad you were able to compile the other tools. Still would
be
> nice to resolve this issue of getting the C++ and Fortran code to
communicate.
> I've searched around and can find no mention of a "-R" option. So I
> have no idea how it's making it into the output of the configure
command.
>
> One possibility is that "-R" is defined somewhere in your
environment
> and its getting passed through by configure. Try running this
command:
> printenv | grep "\-R"
>
> Wherever you find "-R", edit your shell configuration file and
remove it.
>
> And here's another work-around. You could try manually setting
FLIBS
> before running configure. In this settings, I've just switched "-R"
> to "-L".
>
> setenv FLIBS -L${MET_NETCDF}/lib -L${MET_GSL}/lib
> -L/nobackupp8/yliu15/MET/BUFRLIB
> -L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-linux/4.7.3
>
> -L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../.
> ./../x86_64-unknown-linux/lib/../lib64
>
> -L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../.
> ./../lib64
> -L/lib/../lib64 -L/usr/lib/../lib64 -L/nasa/gsl/1.14/lib
> -L/nasa/netcdf/3.6.0/gcc/lib
> -L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../.
> ./../x86_64-unknown-linux/lib
> -L/nasa/pkgsrc/2014Q2/gcc47/lib/gcc/x86_64-unknown-
linux/4.7.3/../../..
> -L/nasa/pkgsrc/2014Q2/gcc47/x86_64-unknown-linux/lib/../lib64
> -L/nasa/pkgsrc/2014Q2/gcc47/lib/../lib64 -lgfortran -lm -lquadmath
> make clean
> ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0
> make install >& make_install.log
>
> Please let me know how it goes.
>
> Thanks,
> John
>
> On Thu, Apr 23, 2015 at 5:05 PM, Liu, Karry at ARB via RT
> <met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
> >
> > Hello John,
> > Yes, I got 17 of these files.
> >
> > Karry
> >
> > -----Original Message-----
> > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > Sent: Thursday, April 23, 2015 3:38 PM
> > To: Liu, Karry at ARB
> > Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
> >
> > Karry,
> >
> > I'd expect 23 minus pb2nc, tc_dland, tc_pairs, tc_stat,
> > modis_regrid, and plot_mode_field which equals 17:
> > ascii2nc*
> > ensemble_stat*
> > gen_circle_mask*
> > gen_poly_mask*
> > grid_stat*
> > madis2nc*
> > mode*
> > mode_analysis*
> > pcp_combine*
> > plot_data_plane*
> > plot_point_obs*
> > point_stat*
> > series_analysis*
> > stat_analysis*
> > wavelet_stat*
> > wwmca_plot*
> > wwmca_regrid*
> >
> > Is that what you're seeing?
> >
> > Thanks,
> > John
> >
> >
> > On Thu, Apr 23, 2015 at 4:31 PM, Liu, Karry at ARB via RT
> > <met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608 >
> > >
> > > Hello John,
> > > I just tried this comment, and it seems compile successfully.
> > > To make sure that it really compiled successfully, how many
> > > executable files I should expect under bin?
> > >
> > > Karry
> > >
> > >
> > > -----Original Message-----
> > > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > > Sent: Thursday, April 23, 2015 3:23 PM
> > > To: Liu, Karry at ARB
> > > Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
> > >
> > > Karry,
> > >
> > > Ah, yes. The same issue is popping when compiling the MET
> > > Tropical Cyclone tools which also link to some Fortran code.
> > > Those can be disabled using "--disable-tc_utils"
> > >
> > > I suspect the following commands would get you a clean compile
> > > that skips over the PB2NC and MET-TC tools:
> > > make clean
> > > ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0
> > > --disable-pb2nc --disable-tc_utils
> > > make install >& make_install.log
> > >
> > > The question remains, how is a "-R" flag making its way into the
> > > FLIBS variable from autoconf? I don't know yet, but will keep
digging.
> > >
> > > Thanks,
> > > John
> > >
> > > On Thu, Apr 23, 2015 at 2:17 PM, Liu, Karry at ARB via RT
> > > <met_help at ucar.edu>
> > > wrote:
> > >
> > > >
> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608
>
> > > >
> > > > Hello John,
> > > > I have recompiled the MET 5.0 and get the same error
message.
> > > > I have attached the log file with this email. Thank you~
> > > >
> > > > Regards,
> > > >
> > > > Karry
> > > >
> > > > -----Original Message-----
> > > > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > > > Sent: Thursday, April 23, 2015 11:55 AM
> > > > To: Liu, Karry at ARB
> > > > Subject: Re: [rt.rap.ucar.edu #71608] fail to compile MET5.0
> > > >
> > > > Karry,
> > > >
> > > > I see that you're having trouble compiling met-5.0. Thanks
for
> > > > sending all those log files. It make it easier to debug.
> > > >
> > > > You're getting a compilation error from the PB2NC tool, which
is
> > > > written in
> > > > C++ but links to some Fortran code. This is an odd error
> > > > C++ message that I've
> > > > never seen before:
> > > > g++: error: unrecognized option '-R'
> > > >
> > > > For some reason, a '-R' flag is being inserted into the
> > > > compilation command for Fortran sources. I've never seen that
> > > > flag before and obviously the
> > > > g++ compiler doesn't like it. I looked at the configure
command
> > > > g++ you used
> > > > and there's nothing odd in there at all.
> > > >
> > > > For now, let's have you compile with the PB2NC tool turned off
> > > > to ensure there aren't other problems. Once everything else
is
> > > > compiling well, we can try to solve this issue.
> > > >
> > > > Try the following:
> > > > make clean
> > > > ./configure --prefix=/nobackupp8/yliu15/MET/met-5.0
> --disable-pb2nc
> > > > make install >& make_install.log
> > > >
> > > > Please let me know how that goes.
> > > >
> > > > Thanks,
> > > > John Halley Gotway
> > > > met_help at ucar.edu
> > > >
> > > >
> > > >
> > > > On Thu, Apr 23, 2015 at 11:27 AM, Liu, Karry at ARB via RT
> > > > <met_help at ucar.edu
> > > > >
> > > > wrote:
> > > >
> > > > >
> > > > > Thu Apr 23 11:27:44 2015: Request 71608 was acted upon.
> > > > > Transaction: Ticket created by Karry.Liu at arb.ca.gov
> > > > > Queue: met_help
> > > > > Subject: fail to compile MET5.0
> > > > > Owner: Nobody
> > > > > Requestors: Karry.Liu at arb.ca.gov
> > > > > Status: new
> > > > > Ticket <URL:
> > > > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=71608
> > > > > >
> > > > >
> > > > >
> > > > > Hello,
> > > > > I was trying to compile MET5.0 using gnu compiler.
> > > > > I used gcc version 4.7.3, netcdf version 3.6.0 and gsl
version
> 1.14.
> > > > > Below is my environment setup:
> > > > > setenv MET_NETCDF /nasa/netcdf/3.6.0/gcc setenv MET_BUFRLIB
> > > > > /nobackupp8/yliu15/MET/BUFRLIB setenv MET_GSL
/nasa/gsl/1.14/
> > > > >
> > > > > When I compiled the MET5.0, I get the error message as
"g++:
> error:
> > > > > unrecognized option '-R'".
> > > > > I have tried different version of gun compiler and all got
the
> > > > > similar error message.
> > > > > I asked IT and they told me that it is very likely a bug in
> > > > > configure file or Makefile rather than compiler issue.
> > > > > I have no idea how to solve this problem. Would you please
> > > > > take a look of four files that I attached with this email
and
> > > > > let me know how can I make this work.
> > > > > Thank you very much!
> > > > >
> > > > > Karry
> > > > >
> > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > >
> > >
> > >
> > >
> >
> >
> >
> >
>
>
>
------------------------------------------------
Subject: fail to compile MET5.0
From: John Halley Gotway
Time: Fri Apr 24 14:27:51 2015
Karry,
So using a newer version of NetCDF resolved this issue you were having
with
pcp_combine. That's good news.
Now I see that we're getting a segmentation fault from the PB2NC tool.
It's odd that your compilation with a NetCDF 3.0 produced a successful
run
of PB2NC while this one with NetCDF 4.0 produces a segmentation fault.
I'm really at a loss here. I have no way of explaining this behavior.
Looks like in my development environment, I'm using NetCDF 4.1.3. I
suppose you could try that, but honestly, the specific version of
NetCDF
people use hasn't come up through MET-Help that often.
John
------------------------------------------------
Subject: fail to compile MET5.0
From: John Halley Gotway
Time: Wed May 20 09:20:23 2015
Just checking in. Have you had any success in compiling the MET
software?
Thanks,
John
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #71608] fail to compile MET5.0
From: Liu, Karry at ARB
Time: Wed May 20 10:39:33 2015
Hello John,
Unfortunately, I still cannot compile PB2NC.
However, I download data from website of DS337.0, which provides
netcdf format data.
So it still works for me. Thank you for asking.
Regards,
Karry
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Wednesday, May 20, 2015 8:20 AM
To: Liu, Karry at ARB
Subject: [rt.rap.ucar.edu #71608] fail to compile MET5.0
Just checking in. Have you had any success in compiling the MET
software?
Thanks,
John
------------------------------------------------
More information about the Met_help
mailing list