[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
Re: NEC SX-4 64 bit IEEE Netcdf
- Subject: Re: NEC SX-4 64 bit IEEE Netcdf
- Date: Tue, 09 Dec 97 12:10:13 -0700
Harvey,
>Date: Tue, 9 Dec 1997 12:20:13 +1100 (EST)
>From: Harvey DAVIES <address@hidden>
>Organization: CSIRO Mathematical and Information Sciences
>To: Steve Emmerson <address@hidden>
>Subject: Re: NEC SX-4 64 bit IEEE Netcdf
>Keywords: 199711262243.PAA12321
In the above message, you wrote:
> Seemed to work OK. Just some warnings like following:
>
> "fort-attio.c", line 12: warning: argument is incompatible with prototype
>
> "fort-lib.c", line 37: warning: improper integer precision : op "="
>
> I can sent the whole file (50853 bytes) if you want it.
I'd like the output from the next try, but only if you get rid of
the "Unvectorized loop" and "... inconsistent with previous reference"
warnings (can you turn them off?).
> Errors occurred. Note apparent infinite loop at end. Output as follows:
>
> + f90 -c -float0 -ew ftest.F
> /stmp/cshld.5467/ow14390_ppdir/i.ftest.F:
>
> f90: warn(84): i.ftest.F, line 2693: Length specification for INTEGER*2 or
> LOGICAL*1 is ignored.
> f90: warn(85): i.ftest.F, line 2697: In block "atts",proper alignment occurred
> because of align option.
> f90: warn(84): i.ftest.F, line 2698: Length specification for INTEGER*2 or
> LOGICAL*1 is ignored.
> f90: vec(3): i.ftest.F, line 2721: Unvectorized loop.
> f90: vec(3): i.ftest.F, line 2728: Unvectorized loop.
> f90: vec(3): i.ftest.F, line 2736: Unvectorized loop.
> f90: warn(62): i.ftest.F, line 2757: Characteristics of arguments of procedure
> "ncagt" are inconsistent with previous
> reference.
> f90: warn(62): i.ftest.F, line 2759: Characteristics of arguments of procedure
> "ncagt" are inconsistent with previous
> reference.
> f90: vec(3): i.ftest.F, line 2768: Unvectorized loop.
> f90: warn(62): i.ftest.F, line 2807: Characteristics of arguments of procedure
> "ncagt" are inconsistent with previous
> reference.
> f90: warn(62): i.ftest.F, line 2814: Characteristics of arguments of procedure
> "ncagt" are inconsistent with previous
> reference.
> f90: i.ftest.F, tncacpy: There are 7 warnings and 4 diagnoses.
>
> f90: vec(3): i.ftest.F, line 4106: Unvectorized loop.
> f90: vec(3): i.ftest.F, line 4117: Unvectorized loop.
> f90: i.ftest.F, tncadel: There are 2 diagnoses.
>
> f90: warn(84): i.ftest.F, line 5406: Length specification for INTEGER*2 or
> LOGICAL*1 is ignored.
> f90: warn(85): i.ftest.F, line 5411: In block "atts",proper alignment occurred
> because of align option.
> f90: warn(84): i.ftest.F, line 5412: Length specification for INTEGER*2 or
> LOGICAL*1 is ignored.
> f90: vec(3): i.ftest.F, line 5435: Unvectorized loop.
> f90: warn(62): i.ftest.F, line 5457: Characteristics of arguments of procedure
> "ncagt" are inconsistent with previous
> reference.
> f90: warn(62): i.ftest.F, line 5459: Characteristics of arguments of procedure
> "ncagt" are inconsistent with previous
> reference.
> f90: vec(3): i.ftest.F, line 5469: Unvectorized loop.
> f90: warn(62): i.ftest.F, line 5508: Characteristics of arguments of procedure
> "ncagt" are inconsistent with previous
> reference.
> f90: warn(62): i.ftest.F, line 5515: Characteristics of arguments of procedure
> "ncagt" are inconsistent with previous
> reference.
> f90: i.ftest.F, tncagt: There are 7 warnings and 2 diagnoses.
>
> f90: warn(84): i.ftest.F, line 6792: Length specification for INTEGER*2 or
> LOGICAL*1 is ignored.
> f90: warn(62): i.ftest.F, line 6837: Characteristics of arguments of procedure
> "ncapt" are inconsistent with previous
> reference.
> f90: warn(62): i.ftest.F, line 6846: Characteristics of arguments of procedure
> "ncapt" are inconsistent with previous
> reference.
> f90: i.ftest.F, tncapt: There are 3 warnings.
>
> f90: warn(85): i.ftest.F, line 6879: In block "atts",proper alignment occurred
> because of align option.
> f90: i.ftest.F, _BLKD: There is 1 warning.
>
> f90: warn(85): i.ftest.F, line 9507: In block "atts",proper alignment occurred
> because of align option.
> f90: vec(3): i.ftest.F, line 9532: Unvectorized loop.
> f90: vec(3): i.ftest.F, line 9557: Unvectorized loop.
> f90: vec(3): i.ftest.F, line 9567: Unvectorized loop.
> f90: vec(3): i.ftest.F, line 9578: Unvectorized loop.
> f90: i.ftest.F, tncinq: There are 1 warning and 4 diagnoses.
>
> f90: warn(85): i.ftest.F, line 10858: In block "atts",proper alignment
> occurred
> because of align option.
> f90: i.ftest.F, tncredf: There is 1 warning.
>
> f90: warn(84): i.ftest.F, line 13488: Length specification for INTEGER*2 or
> LOGICAL*1 is ignored.
> f90: vec(3): i.ftest.F, line 13516: Unvectorized loop.
> f90: vec(3): i.ftest.F, line 13528: Unvectorized loop.
> f90: vec(3): i.ftest.F, line 13540: Unvectorized loop.
> f90: warn(62): i.ftest.F, line 13550: Characteristics of arguments of
> procedure
> "ncvgt" are inconsistent with previous
> reference.
> f90: vec(3): i.ftest.F, line 13555: Unvectorized loop.
> f90: warn(62): i.ftest.F, line 13570: Characteristics of arguments of
> procedure
> "ncvgt" are inconsistent with previous
> reference.
> f90: vec(3): i.ftest.F, line 13575: Unvectorized loop.
> f90: i.ftest.F, tncvgt: There are 3 warnings and 5 diagnoses.
> f90: warn(84): i.ftest.F, line 14877: Length specification for INTEGER*2 or
> LOGICAL*1 is ignored.
> f90: warn(62): i.ftest.F, line 14912: Characteristics of arguments of
> procedure
> "ncvgt1" are inconsistent with previous
> reference.
> f90: warn(62): i.ftest.F, line 14919: Characteristics of arguments of
> procedure
> "ncvgt1" are inconsistent with previous
> reference.
> f90: i.ftest.F, tncvgt1: There are 3 warnings.
>
> f90: warn(84): i.ftest.F, line 16222: Length specification for INTEGER*2 or
> LOGICAL*1 is ignored.
> f90: vec(1): i.ftest.F, line 16247: Vectorized loop.
> f90: vec(3): i.ftest.F, line 16259: Unvectorized loop.
> f90: warn(62): i.ftest.F, line 16267: Characteristics of arguments of
> procedure
> "ncvpt" are inconsistent with previous
> reference.
> f90: vec(3): i.ftest.F, line 16275: Unvectorized loop.
> f90: warn(62): i.ftest.F, line 16283: Characteristics of arguments of
> procedure
> "ncvpt" are inconsistent with previous
> reference.
> f90: i.ftest.F, tncvpt: There are 3 warnings and 3 diagnoses.
>
> f90: warn(84): i.ftest.F, line 17569: Length specification for INTEGER*2 or
> LOGICAL*1 is ignored.
> f90: warn(62): i.ftest.F, line 17598: Characteristics of arguments of
> procedure
> "ncvpt1" are inconsistent with previous
> reference.
> f90: warn(62): i.ftest.F, line 17602: Characteristics of arguments of
> procedure
> "ncvpt1" are inconsistent with previous
> reference.
> f90: i.ftest.F, tncvpt1: There are 3 warnings.
>
> f90: warn(84): i.ftest.F, line 18876: Length specification for INTEGER*2 or
> LOGICAL*1 is ignored.
> f90: warn(62): i.ftest.F, line 18892: Characteristics of arguments of
> procedure
> "ncvgt1" are inconsistent with previous
> reference.
> f90: warn(62): i.ftest.F, line 18893: Characteristics of arguments of
> procedure
> "ncvgt1" are inconsistent with previous
> reference.
> f90: i.ftest.F, tfills: There are 3 warnings.
> ar cru ../libsrc/libnetcdf.a fort-attio.o fort-control.o
> fort-di
> m.o fort-genatt.o fort-geninq.o fort-genvar.o fort-lib.o
> fort-mi
> sc.o fort-v2compat.o fort-vario.o fort-var1io.o fort-varaio.o
> fort-varmio.o fort-varsio.o
> : ../libsrc/libnetcdf.a
> f90 -o ftest -float0 -ew ftest.o ../libsrc/libnetcdf.a
> ./ftest
> *** Testing nccre ...
> *** Testing ncddef ...
> *** Testing ncvdef ...
> *** Testing ncapt, ncaptc ...
> *** Testing ncclos ...
> *** Testing ncvpt1 ...
> *** Testing ncvgt1 ...
> error in ncvgt1 byte: 0 .ne. 122
> error in ncvgt1 short: 0 .ne. 10
> error in ncvgt1 long: 0 .ne. 1000
> error in ncvgt 1 float: 0.000000000000000E+00 not close to
> 3.14159000000000
> 0
> error in ncvgt1 double: 0.000000000000000E+00 not close to
> 0.33333333330000
> 00
> error in ncvg1c
> *** Testing ncvpt ...
> *** Testing ncopn, ncinq, ncdinq, ncvinq, ncanam, ncainq ...
> error in ncinq or ncddef
> error in ncinq or ncvdef
> error in ncinq or ncapt
> error: bad recdim from ncinq
> error in ncainq or ncapt
> error in ncainq or ncapt
> error in ncanam or ncapt
> error in ncainq or ncapt
> error in ncainq or ncapt
>
> This latter message was repeated over 1 million times!!!!
Bummer! I don't know how that could happen.
Apparently, configuring my Alpha's Fortran compiler to use 8-byte
INTEGER's still isn't the same as your SX-4.
Nevertheless, I've got another subdistribution for you. Follow the same
directions as last time.
--------
Steve Emmerson <http://www.unidata.ucar.edu>
