|
|
1.1 root 1: MAME note: this package is derived from the following original SoftFloat
2: package and has been "re-packaged" to work with MAME's conventions and
3: build system. The source files come from bits64/ and bits64/templates
4: in the original distribution as MAME requires a compiler with a 64-bit
5: integer type.
6:
7:
8: Package Overview for SoftFloat Release 2b
9:
10: John R. Hauser
11: 2002 May 27
12:
13:
14: ----------------------------------------------------------------------------
15: Overview
16:
17: SoftFloat is a software implementation of floating-point that conforms to
18: the IEC/IEEE Standard for Binary Floating-Point Arithmetic. SoftFloat is
19: distributed in the form of C source code. Compiling the SoftFloat sources
20: generates two things:
21:
22: -- A SoftFloat object file (typically `softfloat.o') containing the complete
23: set of IEC/IEEE floating-point routines.
24:
25: -- A `timesoftfloat' program for evaluating the speed of the SoftFloat
26: routines. (The SoftFloat module is linked into this program.)
27:
28: The SoftFloat package is documented in four text files:
29:
30: SoftFloat.txt Documentation for using the SoftFloat functions.
31: SoftFloat-source.txt Documentation for compiling SoftFloat.
32: SoftFloat-history.txt History of major changes to SoftFloat.
33: timesoftfloat.txt Documentation for using `timesoftfloat'.
34:
35: Other files in the package comprise the source code for SoftFloat.
36:
37: Please be aware that some work is involved in porting this software to other
38: targets. It is not just a matter of getting `make' to complete without
39: error messages. I would have written the code that way if I could, but
40: there are fundamental differences between systems that can't be hidden.
41: You should not attempt to compile SoftFloat without first reading both
42: `SoftFloat.txt' and `SoftFloat-source.txt'.
43:
44:
45: ----------------------------------------------------------------------------
46: Legal Notice
47:
48: SoftFloat was written by me, John R. Hauser. This work was made possible in
49: part by the International Computer Science Institute, located at Suite 600,
50: 1947 Center Street, Berkeley, California 94704. Funding was partially
51: provided by the National Science Foundation under grant MIP-9311980. The
52: original version of this code was written as part of a project to build
53: a fixed-point vector processor in collaboration with the University of
54: California at Berkeley, overseen by Profs. Nelson Morgan and John Wawrzynek.
55:
56: THIS SOFTWARE IS DISTRIBUTED AS IS, FOR FREE. Although reasonable effort
57: has been made to avoid it, THIS SOFTWARE MAY CONTAIN FAULTS THAT WILL AT
58: TIMES RESULT IN INCORRECT BEHAVIOR. USE OF THIS SOFTWARE IS RESTRICTED TO
59: PERSONS AND ORGANIZATIONS WHO CAN AND WILL TAKE FULL RESPONSIBILITY FOR ALL
60: LOSSES, COSTS, OR OTHER PROBLEMS THEY INCUR DUE TO THE SOFTWARE, AND WHO
61: FURTHERMORE EFFECTIVELY INDEMNIFY JOHN HAUSER AND THE INTERNATIONAL COMPUTER
62: SCIENCE INSTITUTE (possibly via similar legal warning) AGAINST ALL LOSSES,
63: COSTS, OR OTHER PROBLEMS INCURRED BY THEIR CUSTOMERS AND CLIENTS DUE TO THE
64: SOFTWARE.
65:
66: Derivative works are acceptable, even for commercial purposes, provided
67: that the minimal documentation requirements stated in the source code are
68: satisfied.
69:
70:
71: ----------------------------------------------------------------------------
72: Contact Information
73:
74: At the time of this writing, the most up-to-date information about
75: SoftFloat and the latest release can be found at the Web page `http://
76: www.cs.berkeley.edu/~jhauser/arithmetic/SoftFloat.html'.
77:
78:
This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.