This directory contains all the files needed to fix the matlab repmat and multi-dimensional multiply bugs. 

To install:
1) Put all these files in a directory where matlab can see them.
2) There is no step 2. ;-)
   just go ahead and type: repop() or repop_testcases() to check it works. 
   (Actually the codes are written such that matlab will automatically make 
    the required mex files when you first use them -- provided you matlab is 
    set-up correctly this should be transparent to you).

To get help:
help repop
from the matlab prompt


FILELIST

readme           -- this file
mxInfo.c mxInfo.h -- utility functions for storing info about matrices 
repop.c repop.h   -- generic (i.e. non-matlab specific) tprod code
repop_mex.c       -- matlab wrapper code to use repop.c
repop.def         -- macro-code to auto-generate code for the different operators and
                     real/complex argument mixes for repop calls.
ddrepop.c, dsrepop.c, sdrepop.c, ssrepop.c --
                     wrapper source files to generate code for the different 
                     combinationd of double/single inputs. 
repop.m           -- M-file wrapper for the repop help and for the mex-hiding 
                     routines.
repop_testcases.m -- M-file to run a load of tprod unit-tests to verify 
                     correctness.
req.m rge.m rgt.m rldivide.m rle.m rlt.m rminus.m rne.m rplus.m rpower.m rrdivide.m -- 
                     M-file wrappers for the repop help for the different 
							operators

Non-MATLAB specific codes (not needed for MATLAB repop use)
repop_testcase.c  -- non-MATLAB wrapper code to use repop.c
mxUtils.c mxUtils.h -- non-MATLAB helper functions for manipulating mxInfo 
                       structs
valgrind.sh       -- shell-script to run the non-MATLAB code through valgrind


