If anyone has any problems with Alpha itself, the OpenVMS or OSF/1
operating systems, or any of the tools, please contact me and I will
forward it on to the appropriate persons within Digital.
If it is a netCDF software problem, then Unidata will have to solve it
in the distribution kit. This sounds like a problem with the distribution
and how it handles 64 bit longs and ints.
Please let me know what the disposition of this problem is when you find
out.
Thanks.
Rich Lysakowski
Laboratory and R&D Systems Group
Digital Equipment Corporation
======================================================================
From: DECWRL::"montanaro@xxxxxxxxxxxxxxxxxx" "Skip Montanaro" 2-Feb-93 08:07
To: netcdfgroup@xxxxxxxxxxxxxxxx
CC:
Subj: Problems with netCDF on DEC Alpha
We have a DEC Alpha machine in for evaluation. netCDF does not work on it,
apparently because of differences in the assumptions about the sizes of
longs and ints. The output of Steve Pemberton's config program gives:
/* Char = 8 bits, signed */
/* Maximum short = 32767 (= 2**15-1) */
/* Maximum int = 2147483647 (= 2**31-1) */
/* Maximum long = 9223372036854775807 (= 2**63-1) */
/* Alignments for char=1 short=2 int=4 long=8 */
/* Char pointers = 64 bits BEWARE! larger than int! */
/* Int pointers = 64 bits BEWARE! larger than int! */
/* Properties of float: */
/* Base = 2 */
/* Significant base digits = 24 (= at least 7 decimal digits) */
/* Smallest x such that 1.0-base**x != 1.0 = -24 */
/* Small x such that 1.0-x != 1.0 = 5.96046e-08 */
/* Smallest x such that 1.0+base**x != 1.0 = -23 */
/* Smallest x such that 1.0+x != 1.0 = 1.19209e-07 */
/* Arithmetic chops but uses guard digits */
/* Number of bits used for exponent = 8 */
/* Minimum normalised exponent = -125 */
/* Minimum normalised positive number = 1.17549e-38 */
/* The smallest numbers are normalised */
/* Maximum exponent = 128 */
/* Maximum number = 3.40282e+38 */
/* Arithmetic uses a hidden bit */
/* Properties of double: */
/* Base = 2 */
/* Significant base digits = 53 (= at least 15 decimal digits) */
/* Smallest x such that 1.0-base**x != 1.0 = -53 */
/* Small x such that 1.0-x != 1.0 = 1.11022e-16 */
/* Smallest x such that 1.0+base**x != 1.0 = -52 */
/* Smallest x such that 1.0+x != 1.0 = 2.22045e-16 */
/* Arithmetic chops but uses guard digits */
/* Number of bits used for exponent = 11 */
/* Minimum normalised exponent = -1021 */
/* Minimum normalised positive number = 2.22507e-308 */
/* The smallest numbers are normalised */
/* Maximum exponent = 1024 */
/* Maximum number = 1.79769e+308 */
/* Arithmetic uses a hidden bit */
/* Expressions are evaluated in double precision */
/* Memory mallocatable ~= 130976 Kbytes */
Have others run into this problem? If so, is it something that will be fixed
in the "new version" of netCDF?
I haven't been using the Alpha in our group, but can get information on the
breakage from those who have, if it is required.
Thanks,
Skip (montanaro@xxxxxxxxxx)
% ====== Internet headers and postmarks (see DECWRL::GATEWAY.DOC) =====
% Received: by enet-gw.pa.dec.com; id AA24500; Tue, 2 Feb 93 05:09:12 -0800
% Received: by unidata.ucar.edu id AA21846 (5.65c/IDA-1.4.4 for
netcdfgroup-send); Tue, 2 Feb 1993 05:41:18 -070
% Received: from crdgw1.GE.COM by unidata.ucar.edu with SMTP id AA21831
(5.65c/IDA-1.4.4 for <netcdfgroup@xxxxxxxxxxxxxxxx>); Tue, 2 Feb 1993 05:41:13
-070
% Organization: .
% Keywords: 199302021241.AA21831
% Received: by crdgw1.ge.com (5.57/GE 1.145) id AA17051; Tue, 2 Feb 93
07:29:07 ES
% Received: by crdns.crd.ge.com (5.57/sendmail.ease_1.60(1/28/93)) id
AA06656(crdns.crd.ge.com); Tue, 2 Feb 93 07:29:01 -050
% Received: by ausable.crd.Ge.Com (4.1/SMI-4.0/GE-CRD @(#)sun4.ease 1.14
01/06/93) id AA26967; Tue, 2 Feb 93 07:29:00 ES
% Date: Tue, 2 Feb 93 07:29:00 EST
% From: montnaro@xxxxxxxxxxxxxxxxxx (Skip Montanaro)
% Message-Id: <9302021229.AA26967@xxxxxxxxxxxxxxxxxx>
% To: netcdfgroup@xxxxxxxxxxxxxxxx
% Subject: Problems with netCDF on DEC Alpha
% Reply-To: montanaro@xxxxxxxxxxxxxxxxxx (Skip Montanaro)