Frequently Asked Questions
--------------------------

This is a list of questions I received regarding this paper. Hopefully, these
answers will benefit others who are interested in this work.


Q. Is this open-source?
A. No. But, I intend to put up an online demo version of it when I find time.

----

Q. I tried compiling the benchmarks with "-msoft-float" as mentioned in your
   paper, but I get link errors. What am I doing wrong?
A. To successfully compile using '-msoft-float' flag, you need to recompile
   gcc from scratch. Follow these steps:
   1. Download gcc-4.x.x
   2. Add the following lines to gcc-4.2x.x/gcc/config/t-linux
      + # to build the floating point libraries.
      + FPBIT = fp-bit.c
      + DPBIT = dp-bit.c
      +
      + fp-bit.c: $(srcdir)/config/fp-bit.c
      + 	echo '#define FLOAT' > fp-bit.c
      + 	cat $(srcdir)/config/fp-bit.c >> fp-bit.c
      +
      + dp-bit.c: $(srcdir)/config/fp-bit.c
      + 	cat $(srcdir)/config/fp-bit.c >> dp-bit.c
   3. Add '-msoft-float' to all Makefiles that compile DPBIT and FPBIT
      Usually, this is only done in gcc-4.x.x/gcc/mklibgcc.in
      For example, replace "-c $(DPBIT)" with "-c $(DPBIT) -msoft-float"

----