Internet Parallel Computing Archive

Parallel : Acronyms : Cnaps

[ News | IPCA | Mirrors | Add | Search | Mail | Help | WoTUG ]

CNAPS is a registered trademark of Adaptive Solutions, Inc.

Copyright (C) 1994 Adaptive Solutions, Inc.

Permission to use, copy, modify, and distribute this software and its documentation for any purpose and without fee is hereby granted, provided that the above copyright notice and this permission notice appear in supporting documentation, and that the name of Adaptive Solutions, Inc. not be used in advertising or publicity pertaining to distribution of the software without specific, written prior permission. Adaptive Solutions, Inc. makes no representations about the suitability of this software for any purpose. It is provided ``as is'' without express or implied warranty.

• getarg.tar.Z
  Last modified 22:19:00, 22 Mar 1995 - 5.7K - compressed tar
  Header files named getargb07.h and getargb815.h, defining functions which make it possible for CNAPS-C code to to reference "arguments" even though the rules of CNAPS-C are that you have to declare the entry points in the general style entry void main (void) instead of including the "argc" and "argv" parameters which are typically available in ANSI standard and K&R C languages. These functions create assembly-level runtime constants named "CNARG0", "CNARG1", etc., and then make it possible to copy the argument values to an array created in the CNAPS-C code (the arguments should just be integer or character values, not string pointers). For details about which .h file to use and how to call the functions, refer to README.doc in the getarg directory.
• if_xflag.tar.Z
  Last modified 22:19:00, 22 Mar 1995 - 2.9K - compressed tar
  CPL modules that allow implementation of run time conditional branching on the basis of the xflag on _any_ processor being set.
• mat16vec8.tar.Z
  Last modified 22:19:00, 22 Mar 1995 - 4.8K - compressed tar
  an extension to quicklib, supplying a function asiMat16MulVec8 which multiplies a 16-bit matrix by an 8-bit vector, producing a standard 24-bit result vector which is stored like the 32-bit result vector generated by asiMat16MulVec16 (the leftmost 8 bits are just a sign bit extension of the 24-bit element). Although this code may eventually get inserted into matveclib.h and matveclib.sl, at present to access it you just need to #include <mat16mulvec8.h> in your CNAPS-C code, and this header file will issue a CPL command to INDLUDE mat16mulvec8.sl The function asiMat16MulVec8 is called just like asiMat16MulVec16( ... , mat_addr, vec_addr, res_addr) except that, obviously, vec_addr is a "poly char *" pointer instead of a "poly int *" pointer.

  At present, this is only one of two functions available in the relevant directories, although in the future more functions comparable to the matrix-vector multiplies in quicklib, may also be generated -- e.g., asiMat8MulVec16{,File}, and asiMat16MulVec8U to regard the 8-bit vector as an array of unsigned 8-bit integers, in contrast to asiMat16MulVec8 which treats the vector elements as signed 8-bit integers.

• mat16vec8file.tar.Z
  Last modified 22:19:00, 22 Mar 1995 - 4.1K - compressed tar
  a function similar to asiMat16MulVec8, but this one reads the currently open input file as the source of the vector used to multiply the matrix. The relevant #include file for this is called m16v8f.h, and the generated CPL code will contain a line INCLUDE m16v8f.sl The function call is basically identical to asiMat16MulVec16File except that the "file" contains 1-byte objects rather than 16-bit objects to represent the vector elements.
• profile.tar.Z
  Last modified 22:19:00, 22 Mar 1995 - 5.5K - compressed tar
  CPL code for profiling CPL code.
• random.tar.Z
  Last modified 22:19:00, 22 Mar 1995 - 4.6K - compressed tar
  CPL code for parallel pseudo-random number generator.
• seek_input.tar.Z
  Last modified 22:19:00, 22 Mar 1995 - 2.6K - compressed tar
  CPL code for a SEEK_INPUT command, which is functionally the same as the assembler's SEEK command except that it does not require a filename argument (it just defaults to run SEEK on the.
• som.tar.Z
  Last modified 22:19:00, 22 Mar 1995 - 72.7K - compressed tar
  CPL code for self-organizing map.