Configure libzahl
=================

libzahl is configured by editing config.mk. You may choose
to make a copy of config.mk, and reference the copy with
CONFIG when running make. For example:
    cp config.mk my-config.mk
    # edit my-config.mk
    make CONFIG=my-config.mk

Unless you are compiling for Linux you may have to add
    -D'FAST_RANDOM_PATHNAME="<path to a non-blocking random number generator>"'
(/dev/urandom on Linux) and
    -D'SECURE_RANDOM_PATHNAME="<path to a blocking random number generator>"'
(/dev/random on Linux) to CPPFLAGS.

If you are using a C standard library where the higher bits have higher
entropy in the lower bits in rand(3) (as historically was the case),
remove -DGOOD_RAND from CPPFLAGS.

If you don't care if your program crashes on failure, you can add
-DZAHL_UNSAFE to CPPFLAGS. This will give you a marginal performance
boost. You should also add, preferably,
    #define ZAHL_UNSAFE
before including <zahl.h> in your program if you are doing this.

If your CPU does not support indirect jumps (computed jumps) you should
add -DZAHL_ISA_MISSING_INDIRECT_JUMP to CPPFLAGS, and preferably add
    #define ZAHL_ISA_MISSING_INDIRECT_JUMP
before including <zahl.h> in your program.

libzahl contains some (very little) assembly code. In the event
that the used instructions are not supported on your machine, please
report it, and in the meanwhile add -DZAHL_NO_ASM to CPPFLAGS. You
may also have to do this if you are compiling with a compiler that
does not support extended inline assembly. You may also have to add
    #define ZAHL_NO_ASM
to your program before includeing <zahl.h>