1*e790a4ceSJonathan CorbetIntroduction 2*e790a4ceSJonathan Corbet============ 3*e790a4ceSJonathan Corbet 4*e790a4ceSJonathan CorbetThis directory contains the version 0.92 test release of the NetWinder 5*e790a4ceSJonathan CorbetFloating Point Emulator. 6*e790a4ceSJonathan Corbet 7*e790a4ceSJonathan CorbetThe majority of the code was written by me, Scott Bambrough It is 8*e790a4ceSJonathan Corbetwritten in C, with a small number of routines in inline assembler 9*e790a4ceSJonathan Corbetwhere required. It was written quickly, with a goal of implementing a 10*e790a4ceSJonathan Corbetworking version of all the floating point instructions the compiler 11*e790a4ceSJonathan Corbetemits as the first target. I have attempted to be as optimal as 12*e790a4ceSJonathan Corbetpossible, but there remains much room for improvement. 13*e790a4ceSJonathan Corbet 14*e790a4ceSJonathan CorbetI have attempted to make the emulator as portable as possible. One of 15*e790a4ceSJonathan Corbetthe problems is with leading underscores on kernel symbols. Elf 16*e790a4ceSJonathan Corbetkernels have no leading underscores, a.out compiled kernels do. I 17*e790a4ceSJonathan Corbethave attempted to use the C_SYMBOL_NAME macro wherever this may be 18*e790a4ceSJonathan Corbetimportant. 19*e790a4ceSJonathan Corbet 20*e790a4ceSJonathan CorbetAnother choice I made was in the file structure. I have attempted to 21*e790a4ceSJonathan Corbetcontain all operating system specific code in one module (fpmodule.*). 22*e790a4ceSJonathan CorbetAll the other files contain emulator specific code. This should allow 23*e790a4ceSJonathan Corbetothers to port the emulator to NetBSD for instance relatively easily. 24*e790a4ceSJonathan Corbet 25*e790a4ceSJonathan CorbetThe floating point operations are based on SoftFloat Release 2, by 26*e790a4ceSJonathan CorbetJohn Hauser. SoftFloat is a software implementation of floating-point 27*e790a4ceSJonathan Corbetthat conforms to the IEC/IEEE Standard for Binary Floating-point 28*e790a4ceSJonathan CorbetArithmetic. As many as four formats are supported: single precision, 29*e790a4ceSJonathan Corbetdouble precision, extended double precision, and quadruple precision. 30*e790a4ceSJonathan CorbetAll operations required by the standard are implemented, except for 31*e790a4ceSJonathan Corbetconversions to and from decimal. We use only the single precision, 32*e790a4ceSJonathan Corbetdouble precision and extended double precision formats. The port of 33*e790a4ceSJonathan CorbetSoftFloat to the ARM was done by Phil Blundell, based on an earlier 34*e790a4ceSJonathan Corbetport of SoftFloat version 1 by Neil Carson for NetBSD/arm32. 35*e790a4ceSJonathan Corbet 36*e790a4ceSJonathan CorbetThe file README.FPE contains a description of what has been implemented 37*e790a4ceSJonathan Corbetso far in the emulator. The file TODO contains a information on what 38*e790a4ceSJonathan Corbetremains to be done, and other ideas for the emulator. 39*e790a4ceSJonathan Corbet 40*e790a4ceSJonathan CorbetBug reports, comments, suggestions should be directed to me at 41*e790a4ceSJonathan Corbet<scottb@netwinder.org>. General reports of "this program doesn't 42*e790a4ceSJonathan Corbetwork correctly when your emulator is installed" are useful for 43*e790a4ceSJonathan Corbetdetermining that bugs still exist; but are virtually useless when 44*e790a4ceSJonathan Corbetattempting to isolate the problem. Please report them, but don't 45*e790a4ceSJonathan Corbetexpect quick action. Bugs still exist. The problem remains in isolating 46*e790a4ceSJonathan Corbetwhich instruction contains the bug. Small programs illustrating a specific 47*e790a4ceSJonathan Corbetproblem are a godsend. 48*e790a4ceSJonathan Corbet 49*e790a4ceSJonathan CorbetLegal Notices 50*e790a4ceSJonathan Corbet------------- 51*e790a4ceSJonathan Corbet 52*e790a4ceSJonathan CorbetThe NetWinder Floating Point Emulator is free software. Everything Rebel.com 53*e790a4ceSJonathan Corbethas written is provided under the GNU GPL. See the file COPYING for copying 54*e790a4ceSJonathan Corbetconditions. Excluded from the above is the SoftFloat code. John Hauser's 55*e790a4ceSJonathan Corbetlegal notice for SoftFloat is included below. 56*e790a4ceSJonathan Corbet 57*e790a4ceSJonathan Corbet------------------------------------------------------------------------------- 58*e790a4ceSJonathan Corbet 59*e790a4ceSJonathan CorbetSoftFloat Legal Notice 60*e790a4ceSJonathan Corbet 61*e790a4ceSJonathan CorbetSoftFloat was written by John R. Hauser. This work was made possible in 62*e790a4ceSJonathan Corbetpart by the International Computer Science Institute, located at Suite 600, 63*e790a4ceSJonathan Corbet1947 Center Street, Berkeley, California 94704. Funding was partially 64*e790a4ceSJonathan Corbetprovided by the National Science Foundation under grant MIP-9311980. The 65*e790a4ceSJonathan Corbetoriginal version of this code was written as part of a project to build 66*e790a4ceSJonathan Corbeta fixed-point vector processor in collaboration with the University of 67*e790a4ceSJonathan CorbetCalifornia at Berkeley, overseen by Profs. Nelson Morgan and John Wawrzynek. 68*e790a4ceSJonathan Corbet 69*e790a4ceSJonathan CorbetTHIS SOFTWARE IS DISTRIBUTED AS IS, FOR FREE. Although reasonable effort 70*e790a4ceSJonathan Corbethas been made to avoid it, THIS SOFTWARE MAY CONTAIN FAULTS THAT WILL AT 71*e790a4ceSJonathan CorbetTIMES RESULT IN INCORRECT BEHAVIOR. USE OF THIS SOFTWARE IS RESTRICTED TO 72*e790a4ceSJonathan CorbetPERSONS AND ORGANIZATIONS WHO CAN AND WILL TAKE FULL RESPONSIBILITY FOR ANY 73*e790a4ceSJonathan CorbetAND ALL LOSSES, COSTS, OR OTHER PROBLEMS ARISING FROM ITS USE. 74*e790a4ceSJonathan Corbet------------------------------------------------------------------------------- 75