Add compile option to disable floating point usage (#114)

This commit adds the option to disable floating point processing in
QCBOR. To disable, the USEFULBUF_DISABLE_ALL_FLOAT preprocessor macro
needs to be defined.

e.g.:

 $ make CMD_LINE="-DUSEFULBUF_DISABLE_ALL_FLOAT"

This removes the capability (and the code) of decoding floating point
types. The type is still recognised, so a meaningful
QCBOR_ERR_ALL_FLOAT_DISABLED error is returned when a floating point value
is encountered in a decoded qcbor. From the encoding interface the
floating point encoding functions are removed.

Change-Id: I371769246f7d83354607de9bce1e7998b8c536a1
Signed-off-by: Mate Toth-Pal <mate.toth-pal@arm.com>
diff --git a/src/ieee754.c b/src/ieee754.c
index 216cd00..a8079f8 100644
--- a/src/ieee754.c
+++ b/src/ieee754.c
@@ -2,6 +2,7 @@
  ieee754.c -- floating-point conversion between half, double & single-precision
 
  Copyright (c) 2018-2020, Laurence Lundblade. All rights reserved.
+ Copyright (c) 2021, Arm Limited. All rights reserved.
 
  SPDX-License-Identifier: BSD-3-Clause
 
@@ -10,6 +11,12 @@
  Created on 7/23/18
  =============================================================================*/
 
+/*
+ Include before QCBOR_DISABLE_PREFERRED_FLOAT is checked as
+ QCBOR_DISABLE_PREFERRED_FLOAT might be defined in qcbor/qcbor_common.h
+ */
+#include "qcbor/qcbor_common.h"
+
 #ifndef QCBOR_DISABLE_PREFERRED_FLOAT
 
 #include "ieee754.h"