Update CL_FP_FMA definition language to remove ambiguity (#1391)

* Update CL_FP_FMA definition language to remove ambiguity 

Previously, having CL_FP_FMA==0 could be interpreted as the fma builtin function not being implemented. This new wording makes it clearer that the flag only specifies how the builtin function is implemented.

* Fix CL_FP_FMA wording for full-profile.

* Fix CL_FP_FMA wording for half type.

* Fix text layout consistency

Author: paulfradgley

api/embedded_profile.asciidoc

@@ -237,7 +237,8 @@ Device Queries>> table.
         {CL_FP_ROUND_TO_INF} - round to positive and negative infinity
         rounding modes supported
 
-        {CL_FP_FMA} - IEEE754-2008 fused multiply-add is supported.
+        {CL_FP_FMA} - the fma kernel builtin is implemented using
+        hardware-accelerated fused multiply-add as defined in IEEE754-2008.
 
         {CL_FP_CORRECTLY_ROUNDED_DIVIDE_SQRT} - divide and sqrt are correctly
         rounded as defined by the IEEE754 specification.

api/opencl_platform_layer.asciidoc

@@ -948,7 +948,8 @@ include::{generated}/api/version-notes/CL_DEVICE_SINGLE_FP_CONFIG.asciidoc[]
         {CL_FP_ROUND_TO_ZERO_anchor} - round to zero rounding mode supported +
         {CL_FP_ROUND_TO_INF_anchor} - round to positive and negative infinity
         rounding modes supported +
-        {CL_FP_FMA_anchor} - IEEE754-2008 fused multiply-add is supported +
+        {CL_FP_FMA_anchor} - the fma kernel builtin is implemented using
+        hardware-accelerated fused multiply-add as defined in IEEE754-2008. +
         {CL_FP_CORRECTLY_ROUNDED_DIVIDE_SQRT_anchor} - divide and sqrt are correctly
         rounded as defined by the IEEE754 specification +
         {CL_FP_SOFT_FLOAT_anchor} - Basic floating-point operations (such as
@@ -981,7 +982,8 @@ Also see {cl_khr_fp64_EXT}.
         {CL_FP_ROUND_TO_ZERO} - round to zero rounding mode supported +
         {CL_FP_ROUND_TO_INF} - round to positive and negative infinity
         rounding modes supported +
-        {CL_FP_FMA} - IEEE754-2008 fused multiply-add is supported +
+        {CL_FP_FMA} - the fma kernel builtin is implemented using
+        hardware-accelerated fused multiply-add as defined in IEEE754-2008. +
         {CL_FP_SOFT_FLOAT} - Basic floating-point operations (such as
         addition, subtraction, multiplication) are implemented in software
 
@@ -2003,7 +2005,8 @@ include::{generated}/api/version-notes/CL_DEVICE_HALF_FP_CONFIG.asciidoc[]
         {CL_FP_ROUND_TO_ZERO} - round to zero rounding mode supported +
         {CL_FP_ROUND_TO_INF} - round to positive and negative infinity
         rounding modes supported +
-        {CL_FP_FMA} - IEEE754-2008 fused multiply-add is supported +
+        {CL_FP_FMA} - the fma kernel builtin is implemented using
+        hardware-accelerated fused multiply-add as defined in IEEE754-2008. +
         {CL_FP_SOFT_FLOAT} - Basic floating-point operations (such as
         addition, subtraction, multiplication) are implemented in software